Neuropsychologische Mechanismen der Gesichter- und ...

Neuropsychologische Mechanismen der Gesichter- und Körperwahrnehmung Inaugural - Dissertation zur Erlangung des Grades eines Doktors der Naturwissenschaften

in der

Fakultät für Psychologie

der

RUHR - UNIVERSITÄT BOCHUM

vorgelegt von:

Denise Soria Bauser geb. Minnebusch

Gedruckt mit Genehmigung der Fakultät für Psychologie der RUHR-UNIVERSITÄT BOCHUM

Referent:

Denise Soria Bauser

Korreferent:

Prof. Dr. Irene Daum Prof. Dr. Dr. h.c. Onur Güntürkün

Tag der mündlichen Prüfung:

9.07.2009

Neuropsychologische Mechanismen der gesichterund körperwaHRNEHMUNG

Dissertation zur Erlangung des akademischen Grades eines Doktors der Naturwissenschaften (Dr. rer. nat.) in der Fakultät für Psychologie der Ruhr-Universität Bochum

vorgelegt von

Denise Soria Bauser (geborene Minnebusch)

Am Bugapark 156 45899 Gelsenkirchen

DANKSAGUNG Ich bedanke mich bei Prof. Dr. Irene Daum für die Übernahme der Betreuung meiner Arbeit und ihre fachliche Unterstützung in allen Phasen der Entstehung dieser Arbeit. Herrn Prof. Dr. Dr. h.c. Güntürkün möchte ich dafür danken, dass er sich dazu bereit erklärt hat das Koreferat zu übernehmen. Ich danke all meinen Kollegen aus der Abteilung Neuropsychologie für deren Unterstützung und das angenehme Arbeitsklima. Mein besonderer Dank gilt Julia Weiler, Dr. Jennifer Ueckermann und Dr. Boris Suchan für deren Hilfsbereitschaft und fachliche Unterstützung während der letzten drei Jahre. Bei Julia Weiler möchte ich mich noch besonders herzlich für die tolle Arbeitsatmosphäre in unserem Büro und ihre unendliche Geduld beim Korrekturlesen dieser Arbeit bedanken. Meiner Fußballmannschaft möchte ich für die rege Teilnahme an vielen Studien dieser Arbeit danken sowie dafür, dass sie mich während schwieriger Phasen der Entstehung dieser Arbeit aufgemuntert haben. Zuletzt möchte ich mich bei meiner Familie für ihre emotionale Unterstützung während der gesamten Entstehung dieser Arbeit bedanken. Meinen Eltern danke ich dafür, dass sie mich während meiner gesamten Ausbildungszeit bedingungslos unterstützt und stets an mich geglaubt haben. Meinem Mann Ramon möchte ich dafür danken, dass er mich in allen Phasen des Studiums und der Entstehung dieser Arbeit unterstützt und stets ermutigt und aufgemuntert hat.

INHALTSVERZEICHNIS

INHALTSVERZEICHNIS 1.

EINLEITUNG

Seite

6

1.1 1.1.1 1.1.2

Konfigurale Verarbeitungshypothese Verarbeitung menschlicher Gesichter Verarbeitung menschlicher Körperformen

Seite Seite Seite

7 7 9

1.2 1.2.1 1.2.2 1.2.3

Elektrophysiologische Korrelate Elektroenzephalographie (EEG) Verarbeitung menschlicher Gesichter Verarbeitung menschlicher Körperformen

Seite Seite Seite Seite

11 11 11 13

1.3 1.3.1 1.3.2 1.3.3

Neuronanatomische Korrelate Funktionelle Magnetresonanztomographie (fMRT) Verarbeitung menschlicher GesichterSeite 15 Verarbeitung menschlicher KörperformenSeite 16

Seite Seite

15 15

1.4 1.4.1 1.4.2

Selektive Dysfunktionen der Gesichter- und Körperverarbeitung Prosopagnosie Körperformagnosie („body form agnosia“)

Seite Seite Seite

18 18 19

1.5 1.5.1 1.5.2

Modelle der Gesichter- und Körperverarbeitung Verarbeitung menschlicher Gesichter Verarbeitung menschlicher Körperformen

Seite Seite Seite

20 20 21

1.6

Methodische Ansätze zur Überprüfung der Modelle

Seite

23

1.7 1.7.1 1.7.2

Offene Fragen und Zielsetzungen Offene Fragen Ziele der Arbeit

Seite Seite Seite

24 24 25

INHALTSVERZEICHNIS

2.

STUDIEN und ÜBERSICHTSARTIKEL

Seite 28

Publikation 1 Minnebusch DA, Suchan B, Köster O, Daum I. (2009a). A bilateral occipitotemporalc network mediates face perception. Behavioural Brain Research, 198, 179-185.

Seite

28

Publikation 2 Minnebusch DA, Suchan B, Daum I. (2008). Losing your head: behavioral and electrophysiological effects of body inversion. Journal of Cognitive Neuroscience, in Druck.

Seite

36

Publikation 3 Minnebusch DA, Keune PM, Suchan B, Daum I. (2009b). Gradual inversion effects indicate configural processing of human bodies. In Vorbereitung.

Seite

47

Publikation 4 Minnebusch DA, Daum I. (2009a). Are human body forms processed holistically? Evidence from the body composite effect. Cortex, in Begutachtung.

Seite

68

Publikation 5 Minnebusch DA, Daum I. (2009b). Neuronal mechanisms of human body perception. Neuroscience and Biobehavioral Reviews, in Begutachtung

Seite

82

INHALTSVERZEICHNIS

3.

GESAMTDISKUSSION und AUSBLICK

Seite 115

3.1 3.2 3.3 3.3.1 3.3.2 3.4

Neuronale Mechanismen der Gesichterverarbeitung Konfigurale Körperverarbeitung Parallelen und Unterschiede der Gesichter- und Körperverarbeitung Parallelen der Gesichter- und Körperverarbeitung Unterschiede zwischen der Gesichter- und Körperverarbeitung Ausblick

Seite 116 Seite 117 Seite 120 Seite 120 Seite 122 Seite 122

4. ZUSAMMENFASSUNG

Seite 125

5. LITERATURVERZEICHNIS

Seite 126

Curriculum vitae

EINLEITUNG I EINS SEITE 6

1.

EINLEITUNG

Obgleich sich menschliche Gesichter und Körperformen äußerlich sehr deutlich voneinander unterscheiden, weisen sie zahlreiche Gemeinsamkeiten auf, welche für die soziale Kommunikation und Interaktion von besonderer Bedeutung sind. Gesichter und Körperformen erlauben dem Betrachter, andere Personen zu erkennen, und liefern Informationen über deren Alter, Geschlecht, Intentionen und emotionalen Zustand. Die neuronalen Mechanismen der Verarbeitung menschlicher Gesichter stellen seit längerer Zeit einen wichtigen Forschungsschwerpunkt der neurokognitiven Wissenschaft dar (siehe Übersichtsartikel von Kanwisher & Yovel, 2006; Rossion & Gauthier, 2002). Die neuronalen Mechanismen der Körperverarbeitung sind hingegen erst seit einigen Jahren Gegenstand neurowissenschaftlicher Forschungsinteressen (Peelen & Downing, 2007a). Ziel der vorliegenden Arbeit war es, neue Erkenntnisse über die neurokognitiven Verarbeitungsmechanismen menschlicher Gesichter und Körper zu gewinnen. Von besonderem Interesse waren dabei Parallelen und Unterschiede der beteiligten Prozesse und Hirnmechanismen, die beiden Reizklassen zugrunde liegen. Im Zentrum theoretischer Ansätze steht seit einigen Jahren die Hypothese einer konfiguralen Verarbeitung menschlicher Gesichter. Die konfigurale Verarbeitungshypothese postuliert, dass Gesichter primär auf der Grundlage der Konfiguration ihrer Bestandteile erkannt werden, während die Verarbeitung von Gegenständen und anderen Objekten primär auf Einzeldetails beruht. Neuere Befunde deuten darauf hin, dass menschliche Körper, ähnlich wie menschliche Gesichter, konfigural verarbeitet werden. Im Rahmen der Arbeit wurde untersucht, welche Ebenen der konfiguralen Verarbeitung bei der Wahrnehmung von Körperformen von Bedeutung sind. Außerdem wurden die neuronalen Grundlagen der Wahrnehmung menschlicher Gesichter und Körperformen mit Hilfe ereigniskorrelierter Potentiale und funktioneller Magnetresonanztomographie untersucht. Auf der Grundlage dieser Befunde konnten neue Erkenntnisse darüber gewonnen werden, welche neuronalen Prozesse für ein selektives Defizit bei der Gesichterwahrnehmung (Prosopagnosie) verantwortlich sind. In der Einleitung werden in jedem Abschnitt zunächst die bisherigen Befunde zur Gesichterverarbeitung vorgestellt und anschließend die entsprechenden Erkenntnisse zur Körperverarbeitung vorgestellt.


1.1

Konfigurale Verarbeitungshypothese

1.1.1 Verarbeitung menschlicher Gesichter In der aktuellen Literatur ist die Hypothese einer konfiguralen Verarbeitung menschlicher Gesichter von zentraler Bedeutung. Diese geht davon aus, dass Gesichter primär auf der Grundlage der Konfiguration ihrer Bestandteile erkannt werden. Konfigurale Verarbeitung bezieht sich auf Prozesse, welche die Wahrnehmung räumlicher Relationen zwischen einzelnen Bestandteilen eines Reizes beinhaltet (Leder et al., 2001; Leder & Bruce, 2000; Maurer et al., 2002). Die Begriffe konfigurale Verarbeitung und holistische Verarbeitung werden in der Literatur häufig synonym verwendet. Die vorliegende Arbeit orientiert sich an der Definition von Maurer et al. (2002), die von einem Kontinuum der konfiguralen Verarbeitung mit drei Ebenen ausgeht, wobei die holistische Verarbeitung eine Ebene des Kontinuums darstellt (siehe Abbildung 1): Alle menschlichen Gesichter haben die gleichen Basiskonfigurationen bzw. die gleichen „first-order relations“, die sie als Gesichter spezifizieren (z.B. zwei Augen über der Nase). Die holistische Verarbeitung basiert auf der Tatsache, dass die Merkmale eines Gesichts gemeinsam als eine Gestalt wahrgenommen werden. Internale Gesichtsmerkmale sind so stark miteinander verbunden, dass es für den Betrachter sehr schwierig ist, einzelne Gesichtsmerkmale isoliert wahrzunehmen. Die Verarbeitung der „second-order relations“ bezieht sich auf die Distanz zwischen den internalen Gesichtsmerkmalen (z.B. Abstand zwischen beiden Augen). Die Wahrnehmung eines Reizes als Gesicht basiert demnach auf der Verarbeitung der „first-order relations“, wohingegen für die Identifikation einer Person anhand ihres Gesichts die Verarbeitung der „second-order relations“ notwendig ist.

Abbildung 1: Modifizierte Darstellung des konfiguralen Gesichter- und Körperverarbeitungskontinuums und ungeklärte Fragen in Bezug auf die Körperverarbeitung (nach Maurer et al., 2002; Reed et al., 2006).


Als empirischer Beleg einer konfiguralen Verarbeitung wird der Inversionseffekt angesehen (siehe Abbildung 2), d.h. eine verzögerte und weniger genaue Identifikation von Gesichtern bei Drehung um 180 Grad im Vergleich zu aufrecht dargebotenen Gesichtern (Freire et al., 2000; Leder & Bruce, 2000; Yin, 1969; 1970). Der Inversionseffekt wird im experimentellen Kontext in der Regel durch „matching to sample tasks“ untersucht, bei denen zwei Gesichter sukzessiv dargeboten werden. Die Aufgabe der Probanden besteht in diesen Paradigmen darin, so schnell und präzise wie möglich zu entscheiden, ob die beiden Gesichter identisch sind. Der Inversionseffekt ist ein robustes Phänomen, welches in unterschiedlichen Paradigmen nachgewiesen wurde (siehe Übersichtsartikel bei Rossion & Gauthier, 2002). Die meisten Objekte sind invertiert schwieriger zu verarbeiten als bei aufrechter Darbietung. Die Inversion stört jedoch die Verarbeitung menschlicher Gesichter verhältnismäßig stärker als die Verarbeitung anderer Objekte. Alle Ebenen der konfiguralen Verarbeitung sind von dem Inversionseffekt betroffen (Maurer et al., 2002; Reed et al., 2006) a)

b)

c)

d)

Abbildung 2: Darstellung des Inversionseffektes anhand der Thatcher-Illusion (Thompson, 1980). Der groteske Gesichtsausdruck, welcher durch das Rotieren von Augen und Mund entsteht (d), ist bei einem auf dem Kopf stehenden Gesicht (b) kaum mehr zu erkennen. Die Bilder a) und c) zeigen die normale Darstellung (c), sowie die invertierte Version dieser Darstellung (a).

Für die Gesichterwahrnehmung scheinen alle Stufen der konfiguralen Verarbeitung von Bedeutung zu sein. Ein empirischer Beleg für die Bedeutung der „first-order relations“ stammt aus einer Untersuchung von Collishaw und Hole (2000): Vertauscht man die Anordnung der Komponenten eines Reizes, stört dies die Gesichterverarbeitung stärker als die Verarbeitung anderer Reizklassen. Belege für eine konfigurale Verarbeitung von Gesichtern auf holistischer Ebene entstammen Studien zum „face composite effect„ (Abbildung 3): Personen nehmen zwei identische obere Gesichtshälften eher als unterschiedlich wahr, wenn diese mit zwei unterschiedlichen unteren Gesichtshälften dargeboten werden. Das Phänomen tritt auch auf, wenn die Probanden die Instruktion erhalten, die untere Gesichtshälfte zu ignorieren. Dieser Effekt verschwindet jedoch, wenn die obere und untere Gesichtshälfte leicht versetzt dargeboten werden


(Young et al., 1987). Der „face composite effect“ zeigte sich im experimentellen Kontext in „matching to sample tasks“, bei denen die Probanden entscheiden sollen, ob die oberen Gesichtshälften zweier sukzessiv präsentierter Gesichter identisch sind. Die unteren Gesichtshälften beider Stimuli sind stets verschieden (z.B. Goffaux & Rossion, 2006; Hole et al., 1999; Le Grand et al., 2004). Probanden zeigten bessere Leistungen (geringere Fehlerrate und schnellere Reaktionszeiten) bei verschobener und invertierter Darstellung verglichen mit nicht verschobener und aufrechter Reizpräsentation (siehe Abbildung 3). a)

a)

b)

b)

Abbildung 3: Darstellung der „face composite illusion“. Die oberen Gesichtshälften der vier präsentierten Gesichter sind identisch. Paart man diese mit unterschiedlichen unteren Gesichtshälften, so entsteht der Eindruck, dass sie leicht verschieden sind (a). Dieser Effekt verschwindet, wenn die beiden Gesichtshälften leicht versetzt dargeboten werden (b).

1.1.2 Verarbeitung menschlicher Körperformen Der im vorangegangenen Abschnitt beschriebene Inversionseffekt konnte auch für die Wahrnehmung menschlicher Körperformen nachgewiesen werden, was im Sinne einer konfiguralen Verarbeitung interpretiert wurde (Reed et al., 2003; 2006). Probanden zeigten bei invertierten Körperformen schlechtere Leistungen als bei aufrechter Präsentation, was sich in verlangsamten Reaktionszeiten und einer erhöhten Fehlerrate äußerte („matching to sample task“). Des Weiteren existieren empirische Belege für die Bedeutung von „first-order relations“ für die Körperwahrnehmung (Reed et al.,2006). Eine weitere Ebene der konfiguralen Verarbeitung, welche in dem Modell von Maurer et al. (2002) nicht berücksichtigt wurde, bezieht sich auf die hierarchische Struktur eines Reizes (Reed et al., 2006). Für die Wahrnehmung von Körperformen ist nicht nur von Bedeutung, dass sich der Oberkörper oberhalb der


Beine befindet („first-order relations“), sondern auch die genaue Position einzelner Körperteile relativ zu anderen Körperkomponenten (strukturelle Informationen). Die hierarchische Körperstruktur wird beibehalten, auch wenn die Arme und Beine aufgrund von Bewegungen verschiedene Stellungen einnehmen. Diese Art der konfiguralen Verarbeitung befindet sich innerhalb des Kontinuums zwischen der Verarbeitung der „first-order relations“ und der holistischen Verarbeitung. Körperformen scheinen primär anhand dieser räumlichen Beziehungen zwischen den einzelnen Körperteilen identifiziert zu werden. Strukturelle Informationen sind auch für die Gesichterverarbeitung von Bedeutung (Reed et al., 2006; Urgesi et al., 2007). Gesichter und Körper scheinen sich demnach bis zu einem gewissen Grad einen gemeinsamen Verarbeitungspfad zu teilen (siehe Abbildung 1). Bislang fehlen empirische Belege für den Einfluss holistischer Verarbeitung auf die Körperwahrnehmung sowie die Bedeutung von „second-order relations“. Aufgrund dessen könnten, trotz zahlreicher Gemeinsamkeiten, unterschiedliche Ebenen konfiguraler Verarbeitungsprozesse an der Wahrnehmung menschlicher Gesichter und Körperformen beteiligt sein (Reed et al., 2006; Urgesi et al., 2007).


1.2

Elektrophysiologische Korrelate

1.2.1 Elektroenzephalographie (EEG) Die Elektroenzephalographie ist eine nicht-invasive Methode zur Messung der bioelektrischen Hirnaktivität. Der Vorteil dieser Methode liegt, im Vergleich zu anderen bildgebenden Verfahren, in ihrer hohen zeitlichen Auflösung. Die gleichzeitig geringe räumliche Auflösung stellt hingegen einen Nachteil dar. Kognitive Verarbeitungsprozesse führen zu Veränderungen der elektrischen Hirnaktivität. Die dadurch erzeugten elektrischen Potentiale können mit Hilfe der Elektroenzephalographie (EEG) von der Kopfhaut abgeleitet werden. Im spontan auftretenden EEG werden alle Spannungsänderungen des Gehirns abgebildet. In der vorliegenden Arbeit sind vor allem die Veränderungen von Interesse, die nach der Darbietung eines Reizes auftreten, d.h. ereigniskorrelierte Potentiale (EKP).

1.2.2 Verarbeitung menschlicher Gesichter Gesichter evozieren 170 ms nach Präsentation ein negatives EKP mit okzipitotemporalen Maxima (N170). Vergleichbare EKPs werden auch bei der Wahrnehmung anderer Objekte berichtet, allerdings mit einer signifikant geringeren Amplitude (Bentin et al., 1999; Eimer 2000a; b; Minnebusch et al., 2007). Die N170 ist mit der strukturellen Enkodierung eines Reizes assoziiert. In dieser Verarbeitungsphase wird eine globale Gesichtsrepräsentation erstellt, welche für das spätere Wiedererkennen des Gesichts wichtig zu sein scheint (Eimer 2000b; Jacques & Rossion 2007). Die N170 tritt bei der Wahrnehmung invertierter Gesichter verzögert auf. Viele Studien berichten zudem eine größere N170-Amplitude bei der Wahrnehmung invertierter verglichen mit aufrechten Gesichtern (Bentin et al., 1996; Eimer, 2000a; Itier & Taylor, 2004a; b; Marzi & Viggiano, 2007; Rossion et al., 1999; 2000b).


Abbildung 4: Exemplarische Darstellung der P100 und N170 bei der Wahrnehmung von Körpern, Gesichtern und Objekten (nicht publizierte Daten).

Neben der N170 haben Gesichter einen Einfluss auf ein frühes positives endogenes EKP, welches 100-120 ms nach der Reizdarbietung auftritt (P100). Die P100 weist eine größere Amplitude bei der Verarbeitung von Gesichtern vergleichen mit anderen Objekten auf sowie eine verzögerte Latenz und größere Amplitude für invertierte verglichen mit aufrechten Gesichtern (Itier & Taylor, 2004a; b; Rossion et al.,1999; 2000b). Zudem scheint die P100 durch Prozesse generiert zu werden, welche mit der Wahrnehmung eines Reizes als Gesicht assoziiert sind (Herrmann et al., 2005; Itier & Taylor, 2004b). Quellenlokalisationen identifizierten den Ursprung der P100 im fusiformen Gyrus sowie den Ursprung der N170 im parieto-temporalen Netzwerk, welches den fusiformen Gyrus sowie den lateralen temporalen Kortex beinhaltet (Herrmann et al., 2005; Pizzagalli et al., 2002).


Abbildung 5: Exemplarische Darstellung des elektrophysiologischen Inversionseffektes (nicht publizierte Daten).

Der Inversionseffekt wird als Indikator für konfigurale Verarbeitungsprozesse angesehen und scheint sowohl die P100 als auch die N170 zu modulieren. Unklar ist, zu welchem Zeitpunkt die konfigurale Verarbeitung stattfindet. Zudem ist fraglich, ob der behaviorale Inversionseffekt ab einem gewissen Rotationswinkel auftritt oder ein linearer bzw. quadratischer Zusammenhang zwischen dem Grad der Rotation und dem Grad der Leistungsbeeinträchtigung der Probanden besteht. Jacques und Rossion (2007) zeigten ihren Probanden Gesichter, welche in Schritten von 30°, von der aufrechten Darbietung ausgehend, rotiert waren. Die Aufgabe der Probanden bestand darin zu entscheiden, ob zwei sukzessiv dargebotene Reize identisch waren oder nicht („matching to sample task“). Ziel der Studie war es, sowohl die P100 und N170 als auch das Verhaltensmuster der Probanden systematisch zu variieren. Die Autoren fanden einen quadratischen Zusammenhang zwischen dem Rotationswinkel und den Verhaltensdaten sowie dem Rotationswinkel und den elektrophysiologischen Befunden (P100, N170). Allerdings korrelierten die Verhaltensdaten nur mit der N170, aber nicht mit der frühen P100, was darauf hindeutet, dass der behaviorale Inversionseffekt erst während der perzeptuellen Enkodierung eines Gesichts auftritt.

1.2.2 Verarbeitung menschlicher Körperformen Körper evozieren eine P100, welche eine größere Amplitude bei der Wahrnehmung von Körpern verglichen mit anderen Objekten zeigt (Thierry et al., 2006; siehe Abbildung 4). Analog zur N170 wurde ein weiteres EKP in Zusammenhang mit der Wahrnehmung von Körperformen beschrieben (Thierry et al., 2006). Das körperspezifische Potential tritt 170-190 ms (N190) nach der Reizpräsentation auf und seine


Quelle unterscheidet sich im rechten posterioren extrastiatären Kortex von der Quelle der gesichtsspezifischen N170 (Kovacs et al., 2006; Stekelenburg & de Gelder, 2004; Thierry et al., 2006). Invertierte Körper lösen eine signifikant verzögerte und größere N190 aus als aufrecht dargebotene Körper (Righart & de Gelder, 2007; siehe Abbildung 5). Diese Effekte unterstützen die Annahme einer konfiguralen Körperverarbeitung. In den bislang einzigen EEG-Studien, die den Inversionseffekt bei Körpern untersucht haben (Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004), minimierten die Autoren den Einfluss gesichtsspezifischer Verarbeitungsmechanismen, indem sie Körper mit maskierten Gesichtern als Stimulusmaterial präsentierten. Studien mittels funktioneller Magnetresonanztomographie (fMRT) weisen jedoch darauf hin, dass kontextuelle Hinweise (z.B. maskierte Gesichter) stimulusspezifische Aktivierungen auslösen können (Cox et al., 2004). Zudem wurde belegt, dass körperspezifische Aktivierungen in okzipitotemporalen Arealen durch die Darbietung des Gesichts moduliert werden (Morris et al., 2006). Aufgrund dessen ist bislang unklar, ob die berichteten Effekte auf die Körperform, das Gesicht oder beide Reize zurückzuführen sind. Des Weiteren ist fraglich, ob es sich bei der N190 um eine Komponente handelt, welche auf die Wahrnehmung menschlicher Körper spezialisiert ist, oder ob die N190 auf Körper allgemein (auch Tiere) generalisiert.


1.3

Neuroanatomische Korrelate

1.3.1 Funktionelle Magnetresonanztomographie (fMRT) Die Magnetresonanztomographie (MRT) ist eine nicht-invasive Technik zur dreidimensionalen Darstellung innerer Organe, Gewebe und Gelenke mittels Magnetfeldern und hochfrequenten Signalen im Radiowellenbereich. Im Gegensatz zum EEG weist die MRT eine hohe räumliche und eine geringe zeitliche Auflösung auf.

1.3.2 Verarbeitung menschlicher Gesichter Die neuroanatomische Basis der Gesichterverarbeitung wird vorrangig in einer Region des fusiformen Gyrus vermutet, welche als „fusiform face area“ (FFA) bezeichnet wird (Kanwisher et al., 1997; Kanwisher & Yovel, 2006; siehe Abbildung 6). Diese Annahme beruht auf der Beobachtung, dass diese Region bei der Wahrnehmung menschlicher Gesichter mit einem stärkeren Aktivierungsanstieg reagiert als bei der Wahrnehmung anderer Objekte. Gesichtsspezifische Aktivierungen innerhalb der FFA sind in der rechten Hemisphäre stärker ausgeprägt als in der linken (siehe Übersichtsartikel von Kanwisher & Yovel, 2006). Der Aktivierungsanstieg innerhalb der FFA ist zudem größer für aufrecht dargebotene verglichen mit invertierten Gesichtern (Kanwisher et al., 1998; Yovel & Kanwisher, 2004). Dieser Effekt korreliert mit dem behavioralen Gesichterinversionseffekt (Yovel & Kanwisher, 2005), was darauf hindeutet, dass Gesichter innerhalb der FFA konfigural verarbeitet werden und dass dieses Areal ein mögliches neuronales Korrelat des Inversionseffektes darstellt. Obgleich angenommen wird, dass die FFA an der Wahrnehmung und Identifikation von Gesichtern beteiligt ist, herrscht in der Literatur keine Einigkeit über ihre genaue Funktion (Grill-Spector et al., 2004; Kanwisher et al., 1997; Kanwisher & Yovel, 2006). Die meisten Autoren vertreten die Annahme, dass es sich bei der FFA um ein gesichtssensitives Areal handelt, obwohl signifikante FFA-Aktivierungen auch bei der Wahrnehmung anderer Objekte nachgewiesen werden konnten (Haxby et al., 2001; Ishai et al., 1999; 2000). Fraglich ist, ob die FFA hinreichend oder notwendig für die erfolgreiche Gesichterverarbeitung ist. Zusätzlich zur FFA lösen Gesichter größere Aktivierungen im superioren temporalen Sulkus (STS) und inferioren okzipitalen Gyrus aus (Abbildung 6), wiederum mit einer Dominanz der rechten Hemisphäre (Haxby et al., 1999; Kanwisher et al., 1997). Die gesichtssensitive Region im okzipitalen Gyrus wird als „occipital face area“ (OFA) bezeichnet und verarbeitet die physikalischen Merkmale des Gesichts, während der STS mit der Verarbeitung dynamischer Gesichtsmerkmale assoziiert ist (z.B. Gesichtsausdruck).


Abbildung 6: Darstellung der gesichtssensitiven „fusiform face area“ (FFA) und „occipital face area“ (OFA) sowie der körpersensitiven „extrastriate body area“ (EBA) und „fusiform body area“ (FBA; Taylor et al., 2007)

1.3.3 Verarbeitung menschlicher Körperformen Studien mittels funktioneller Bildgebung haben zwei Regionen im extrastriatären visuellen Kortex identifiziert, die auf die Wahrnehmung menschlicher Körperformen spezialisiert zu sein scheinen (Downing et al., 2001; 2006; Peelen & Downing, 2005; Schwarzlose et al., 2005; Abbildung 6). Die extrastriatäre Körperregion („extrastriate body area“; EBA) zeigt größere Aktivierungen bei der Wahrnehmung menschlicher Körperformen und Körperteile im Vergleich zu anderen Reizen (Downing et al., 2001; Urgesi et al., 2004). EBA-Aktivierungen wurden zudem bei der Wahrnehmung von Strichzeichnungen eines Körpers und von Körpersilhouetten berichtet. Demnach scheint die EBA über eine abstrakte Körperrepräsentation zu verfügen und auf die Identifikation eines Reizes als menschlichen Körper spezialisiert zu sein (Downing et al., 2001). Zudem wurden größere EBA-Aktivierungen für Körper mit Kopf verglichen mit Körpern ohne Kopf gefunden, was darauf hindeutet, dass der Darbietung des Kopfes eine besondere Bedeutung bei der Körperverarbeitung zukommt (Morris et al., 2006). Die EBA befindet sich im posterioren Ende des inferioren okzipitalen Sulkus (Abbildung 6) und überlappt teilweise mit einer bewegungssensitiven Region („motion sensitive area“; hMT) und einem Areal, welches auf die Verarbeitung von Objektformen spezialisiert ist (lateral occipital complex; LOC; Downing et al., 2007; Peelen & Downing, 2007a). Studien mittels „multi-voxel pattern analysis“ (MVPA) deuten jedoch auf eine funktionelle Unabhängigkeit dieser Regionen (EBA, hMT und LOC) hin (Downing et al., 2007; Peelen & Downing, 2007b). Die MVPA fokussiert nicht auf individuelle Voxel (wie das traditionelle allgemeine lineare Modell), sondern verwendet bestimmte Algorithmen, um Aktivierungsmuster zu


klassifizieren. Ziel dieser Analyse ist die Identifikation der Informationen, die sich in einem gegebenen Aktivierungsmuster befinden (Norman et al., 2006). Die im Gyrus fusiformis gelegene fusiforme Körperregion („fusiform body area;“ FBA) wird im Unterschied zur EBA durch die Wahrnehmung des gesamten Körpers und nicht durch die Wahrnehmung einzelner Körperteile aktiviert (Peelen et al., 2006; Peelen & Downing, 2005; Schwarzlose et al., 2005). Zudem überlappt die FBA teilweise mit der gesichtssensitiven FFA (Peelen et al., 2006; Peelen & Downing, 2005; Schwarzlose et al., 2005).


1.4

Selektive Dysfunktionen der Gesichter- und Körperverarbeitung

1.4.1 Prosopagnosie Für die Spezifität der Verarbeitung von Gesichtern sprechen Beobachtungen einer relativ umschriebenen Beeinträchtigung des Erkennens bekannter Personen anhand von Gesichtern (Prosopagnosie; Bodamer, 1947). Man unterscheidet zwischen einer erworbenen und angeborenen Variante der Prosopagnosie. Die erworbene Prosopagnosie („acquired prosopagnosia“) tritt vor allem nach Schädigungen des rechten okzipitotemporalen Kortex auf, vorrangig innerhalb der FFA und/oder OFA. Untersuchungen an Probanden mit erworbener Prosopagnosie deuten darauf hin, dass die funktionelle Integrität eines bilateralen Netzwerks, in das FFA und OFA eingebunden sind, für eine intakte Gesichterverarbeitung notwendig ist (Barton et al., 2002; Rossion et al., 2003; Sorger et al., 2007; Steeves et al., 2006; Wada & Yamamoto, 2001). Bei der angeborenen Variante unterscheidet man zudem zwischen Prosopagnostikern, bei denen zu keinem Entwicklungszeitpunkt Hinweise auf Hirnschädigungen bestehen („congenital prosopagnosia“), und jenen, bei denen vor der Geburt oder während der frühen Entwicklungsphasen Hirnschädigungen aufgetreten sein könnten („developmental prosopagnosia“; Barton et al., 2003; Behrmann et al., 2005; Behrmann & Avidan, 2005). Bei der angeborenen Prosopagnosie handelt es sich um eine sehr seltene Erkrankung und die Ursache sowie die zugrundeliegenden Mechanismen sind bislang unklar. Es scheint sich hierbei um ein heterogenes Störungsbild zu handeln, welches nicht auf Defizite innerhalb eines spezifischen Aspektes der Gesichterverarbeitung zurückgeführt werden kann (Le Grand et al., 2006). Die Annahme eines heterogenen Verarbeitungsdefizits wird durch elektrophysiologische (Harris et al., 2005; Minnebusch et al., 2007) und fMRT-Studien (Avidan et al., 2005; Bentin et al., 2007; Hadjikhani & de Gelder, 2002; Hasson et al., 2003) an gesunden Kontrollprobanden und Personen mit angeborener Prosopagnosie unterstützt. Die neuronalen Korrelate der Gesichter- und Körperverarbeitung liegen in benachbarten und zum Teil überlappenden kortikalen Regionen (Peelen & Downing, 2007a; Schwarzlose et al., 2005). Gesichter und Körper lösen vergleichbare EKPs aus. Zudem scheinen konfigurale Verarbeitungsmechanismen für beide Reizklassen von zentraler Bedeutung zu sein. Righard und de Gelder (2007) konnten zeigen, dass Gesichter und Körper bei Prosopagnostikern, verglichen mit gesunden Kontrollprobanden, veränderte EKPs evozieren. Demnach könnten Gesichter und Körper konfigurale Verarbeitungsmechanismen aktivieren, welche in frühen Enkodierungsphasen von Bedeutung sind.


1.4.2 Körperformagnosie („body form agnosia“) Bisher gibt es keine Erkenntnisse über ein körperspezifisches Verarbeitungsdefizit analog der Prosopagnosie (Peelen & Downing, 2007a). Das Fehlen dieses Defizits stellt bislang den größten Unterschied zwischen den beiden Reizklassen dar. In einer kürzlich erschienenen Studie berichten Moro et al., (2008) Zusammenhänge zwischen spezifischen Verarbeitungsdefiziten und umschriebenen Läsionsmustern. Läsionen, welche die EBA umfassen, waren mit einer Körperformagnosie („body form agnosia“) assoziiert, d.h. mit einem selektiven Defizit bei der Verarbeitung von Körperformen, und Läsionen im ventralen prämotorischen Kortex mit einer „body action agnosia“, d.h. Schwierigkeiten bei der Wahrnehmung von Körperbewegungen. Die Körperformagnosie wurde in dieser Studie jedoch nur mittels Tests zur Verarbeitung von Körperteilen, einzelnen Gesichtsmerkmalen (z.B. Auge, Nase) und Objektteilen untersucht („matching to sample task“). Für eine detaillierte Diagnose der postulierten Defizite sollte hingegen die Verarbeitung ganzer Körperformen und Gesichter untersucht werden. Des Weiteren scheint ein Vergleich mit Personen mit erworbener Prosopagnosie notwendig zu sein, um Aussagen über die Existenz einer Körperformagnosie treffen zu können, welche sich von der Prosopagnosie unterscheidet. Aufgrund dessen kann derzeit nicht von einem der Prosopagnosie vergleichbaren Defizit der Verarbeitung menschlicher Körperformen ausgegangen werden.


1.5

Modell der Gesichter- und Körperverarbeitung

1.5.1 Verarbeitung menschlicher Gesichter Haxby et al. (2000) integrierten die vorliegenden empirischen Befunde zur Wahrnehmung menschlicher Gesichter in ein hierarchisches Modell mit drei Ebenen (vgl. Abbildung 7): Diese Ebenen umfassen den inferioren Teil des Gyrus okzipitalis (entspricht der OFA), den lateralen Gyrus fusiformis (oder FFA) und den STS. Die OFA analysiert individuelle Gesichtsmerkmale, während die FFA invariante, identitätsspezifische Aspekte des Gesichts verarbeitet. Der STS ist auf die Verarbeitung veränderbarer Gesichtsmerkmale (z.B. Gesichtsausdruck, Lippenbewegungen) spezialisiert. Es existieren Hinweise auf konfigurale Verarbeitungsprozesse in der FFA, nicht aber in der OFA (Yovel & Kanwisher, 2005).

Abbildung 7: Modell der kortikalen Gesichter- und Körperverarbeitung. Der Pfad für die Gesichterverarbeitung orientiert sich an dem Modell von Haxby et al. (2000).

Gesichtsspezifische Informationen werden von der OFA zur FFA und zum STS weitergeleitet (Fairhall & Ishai, 2007; Haxby et al., 2000). Die rechte OFA soll zudem reziprok mit der rechten FFA verbunden sein, wobei die funktionelle Integrität beider Regionen für eine intakte Verarbeitung von Gesichtern notwendig ist (Rossion et al., 2003). Rossion (2008) beschreibt zudem einen direkten Verbindungsweg vom primären visuellen Kortex zur FFA. Zusammenfassend bleibt festzuhalten, dass Gesichter ein weit verbreitetes kortikales Netzwerk aktivieren. Die gemeinsame Aktivität dieses Netzwerkes scheint für die erfolgreiche Gesichterverarbeitung von Bedeutung zu sein.


1.5.2 Verarbeitung menschlicher Körperformen Bislang wurden zwei Modelle der Körperverarbeitung postuliert. Das erste Modell basiert auf Gemeinsamkeiten zwischen der Gesichter- und Körperverarbeitung und konzentriert sich primär auf die EBA und FBA (Taylor et al., 2007). Das zweite Modell geht davon aus, dass Körper ein weit verbreitetes kortikales Netzwerk aktivieren (Urgesi et al., 2004; 2007). Analog zu dem Gesichterverarbeitungsmodell (Haxby et al., 2000) lassen sich bezüglich der Körperverarbeitung folgende Hypothesen formulieren: Visuelle Informationen über Körperformen werden zunächst in der EBA auf der Ebene einzelner Details verarbeitet (siehe Abbildung 7). Von dort gelangen diese Informationen in die FBA, in der die einzelnen Körperteile zu einem Gesamtbild integriert werden. Taylor et al. (2007) untersuchten die Aktivierung in der EBA und FBA in einer Studie, in der sie die Menge an relevanten Informationen systematisch variierten (die Darbietung begann mit der Präsentation eines Fingers und endete bei der Präsentation des gesamten Körpers). Die EBA zeigte einen linearen Aktivierungsanstieg in Abhängigkeit von der Anzahl der Körperteile, die gezeigt wurden. In der FBA stieg die Aktivierung sprunghaft an, sobald der gesamte Körper gezeigt wurde, während keine spezifische Aktivierung bei der Wahrnehmung einzelner Finger oder Hände nachgewiesen werden konnte. Allerdings liegen bislang keine Studien vor, die die funktionelle Interaktion zwischen der EBA und FBA untersucht haben.

Abbildung 8: Modell der Körperverarbeitung nach Urgesi et al. (2007).


Das zweite Modell nimmt die Existenz zweier unabhängiger Verarbeitungspfade an, welche auf die Wahrnehmung menschlicher Körperformen spezialisiert sind (Urgesi et al., 2004; 2007; siehe Abbildung 8). Die Funktion der FBA wird in diesem Modell nicht erläutert. Das Modell orientiert sich an der Einteilung in einen dorsalen und einen ventralen visuellen Verarbeitungspfad (Goodale & Milner, 1992; Ungerleider & Haxby, 1994). Die erste Route besteht aus Arealen des dorsalen Verarbeitungspfades. Innerhalb dieses Pfades werden beobachtete Körperhaltungen anderer Personen mit der sensomotorischen Repräsentation des eigenen Körpers in Verbindung gesetzt, was auf einer konfiguralen Verarbeitung des gesamten Körpers beruht. In diesem Modell bezieht sich die konfigurale Verarbeitung auf die sensomotorische Körperrepräsentation (Gyrus postzentralis im Parietallappen) und nicht auf die visuelle Darstellung im extrastriatären visuellen Kortex. Der zweite Pfad ist auf die Verarbeitung lokaler Körpermerkmale spezialisiert (z.B. Körperteile, Körperform) und ist mit der EBA assoziiert.


1.6

Methodische Ansätze zur Überprüfung der Modelle

Bildgebende Studien deuten darauf hin, dass die FFA bei der Gesichterverarbeitung von besonderer Bedeutung ist (siehe 1.3). Die Beobachtung normaler FFA-Aktivierungen bei Prosopagnosie deutet jedoch darauf hin, dass diese nicht hinreichend für eine intakte Gesichterverarbeitung sind (Avidan et al., 2005; Hasson et al., 2003). Aufgrund dessen stellt sich die Frage, wie gesichtssensitive Hirnareale miteinander interagieren. Im Folgenden werden zwei neuere methodische Ansätze vorgestellt, mittels derer diese Interaktionen untersucht werden können. Neuere Befunde an gesunden Kontrollprobanden mittels „dynamic causal modeling“ (DCM) zeigten, dass ein Netzwerk bestehend aus FFA und OFA der Gesichterverarbeitung zugrunde liegt (Fairhall & Ishai, 2007; Ishai, 2008). DCM stellt einen Analyseansatz dar, mit dessen Hilfe die effektive Konnektivität innerhalb eines Netzwerkes analysiert werden kann (Friston et al., 2003). Die Analyse bezieht sich auf experimentell induzierte Veränderungen der effektiven Konnektivität zuvor definierter Hirnareale, unter Berücksichtigung ihrer nichtlinearen und dynamischen Eigenschaften. Ziel dieser Analyse ist, Aussagen über die Interaktion kortikaler Areale treffen zu können. DCM-Analysen setzen demnach Annahmen über beteiligte Hirnareale voraus. Es besteht die Möglichkeit, dass das gesichtssensitive kortikale Netzwerk aus weiteren Arealen besteht, deren Funktion noch nicht hinreichend bekannt ist. Diese Areale können mit DCM-Analysen nicht identifiziert werden. Die psychophysiologische Interaktionsanalyse (PPI) stellt eine weitere Analysemethode dar (Friston et al., 1997), mittels derer Regionen im Gehirn identifiziert werden können, die während eines bestimmten Verarbeitungsprozesses (Gesichterverarbeitung) eine starke Interaktion mit einer zuvor definierten Region (z.B. FFA) aufweisen und eine schwache Interaktion während eines anderen Verarbeitungsprozesses (z.B. Objektverarbeitung). Ihr Vorteil gegenüber der DCM-Analyse liegt darin, dass sie, ausgehend von einer vorher definierten Region, kortikale Areale identifiziert, welche bei der Gesichterverarbeitung vergleichbare Aktivierungsänderungen zeigen. Psychophysiologische Interaktion bedeutet, dass sich die Aktivierung einer Region und die CoAktivierungen anderer an den Prozessen beteiligter Areale durch den experimentellen oder psychologischen Kontext signifikant verändert. Mit Hilfe der PPI-Analyse kann die kortikale Aktivität einer bestimmten Region durch die Interaktionen experimenteller Parameter mit anderen Arealen erklärt werden. Psychophysiologische Interaktion bezieht sich demnach in diesem Zusammenhang darauf, dass die Aktivierungsstärke einer bestimmten Region auf der Basis anderer kortikaler Aktivitäten und experimenteller Variationen vorhergesagt werden kann. Bislang liegen keine PPI-Studien vor, welche die Interaktionen zwischen den Arealen des gesichtssensitiven Netzwerkes untersuchen. Während der Gesichterverarbeitung scheint die rechte FFA von besonderer Bedeutung zu sein, obgleich FFA-Aktivierungen für die erfolgreiche Gesichterverarbeitung nicht ausreichend sind (siehe 1.3.2). Aufgrund dessen könnten PPI-Analysen an gesunden Probanden Hinweise darauf liefern, wie die gesichtssensitiven Areale miteinander interagieren.


1.7

Offene Fragen und Ziele der Arbeit

1.7.1 Offene Fragen Bisherige Befunde deuten darauf hin, dass es sich bei der angeborenen Prosopagnosie um ein heterogenes Verarbeitungsdefizit handelt (Avidan et al., 2005; Harris et al., 2005; Le Grand et al., 2006; Minnebusch et al., 2007). Die konträre Befundlage ist ein Grund dafür, dass die genaue Ursache der angeborenen Prosopagnosie bislang unklar ist. Studien an gesunden Kontrollprobanden deuten darauf hin, dass der rechten FFA bei der Gesichterverarbeitung eine zentrale Bedeutung zukommt (siehe 1.3.2). Bildgebungsstudien bei Probanden mit einer angeborene Prosopagnosie berichten normale Aktivierungsmuster in der FFA (z.B. Avidan et al., 2005; Hasson et al., 2003). Demzufolge scheinen FFA-Aktivierungen per se nicht ausreichend für eine erfolgreiche Gesichterverarbeitung zu sein. Analysemethoden wie DCM und PPI, mittels derer Interaktionen gesichtsspezifischer kortikaler Areale während der Gesichterverarbeitung untersucht werden können, könnten Hinweise auf mögliche Ursachen der angeborenen Prosop-agnosie liefern. Belege für die Bedeutung der FFA und OFA entstammen vorherigen DCM-Analysen (Fairhall & Ishai, 2007), wobei diese Art der Datenauswertung an strenge Vorannahmen gebunden ist (siehe 1.6). Vergleichbare PPI-Analysen, welche ohne Vorannahmen über die beteiligten Areale auskommen, könnten weitere gesichtsspezifische Areale identifizieren, welche Teil des Gesichterverarbeitungsnetzwerkes sind. Zahlreiche Studien haben sich mit den neuronalen Mechanismen der Gesichterverarbeitung beschäftigt. Einige neuere Befunde deuten auf Gemeinsamkeiten bei der Gesichter- und Körperverarbeitung hin, wobei die geringe Anzahl der bisherigen Studien bislang keine eindeutige Aussage über potentielle Parallelen und Unterschiede zulässt. Die stärksten Hinweise darauf, dass an der Verarbeitung von Gesichtern und Körpern vergleichbare Mechanismen beteiligt sind, liefern Studien, die für beide Stimulusklassen einen elektrophysiologischen Inversionseffekt nachweisen (Bentin et al., 1996; Eimer, 2000a; Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004; siehe 1.1 und 1.2). Der Inversionseffekt ist ein Beleg dafür, dass Gesichter (und vermutlich auch Körper) anders verarbeitet werden als andere Objekte, was durch die konfigurale Verarbeitungshypothese erklärt werden kann. Die wenigen EEG-Studien, die diesen Effekt bei der Körperverarbeitung untersucht haben, präsentierten Körper mit maskierten Gesichtern als Stimuli (Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004). Aufgrund dessen ist bislang unklar, ob der berichtete elektrophysiologische Körperinversionseffekt durch die Körperform, das maskierte Gesicht oder beides ausgelöst wurde. Des Weiteren besteht die Möglichkeit einer Generalisierung der beobachteten Effekte im Sinne eines Inversionseffektes für Körperformen allgemein (menschliche Körperformen und Tierkörper), was bedeuten würde, dass der Inversionseffekt nicht spezifisch für menschliche Körperformen ist.


Die Entstehung des Körperinversionseffektes wurde bislang weder in Verhaltens- noch in elektrophysiologischen Studien untersucht. Die Frage, welchen Einfluss der Grad der Rotation des Körperstimulus auf die Körperverarbeitung hat, steht in diesem Zusammenhang im Fokus des Interesses. Für die Verarbeitung von Gesichtern wurde ein quadratischer Zusammenhang berichtet (Jacques & Rossion, 2007). Des Weiteren ist bislang unklar, zu welchem Zeitpunkt der Körperinversionseffekt auftritt. Analog zu den Befunden zur Gesichterverarbeitung (Jacques & Rossion, 2007) kann angenommen werden, dass die konfigurale Körperverarbeitung während der strukturellen Enkodierung des Reizes auftritt, d.h. 170 ms nach der Stimuluspräsentation. Maurer et al. (2002) postulieren ein Kontinuum der konfiguralen Gesichterverarbeitung, welches aus den folgenden drei Ebenen besteht: „first-order relations“, holistische Verarbeitung und „second-order relations“ (siehe 1.1.1). Neuere theoretische Konzepte gehen mit der Fokussierung auf die hierarchische Struktur eines Reizes von einer weiteren Ebene aus, welche sich zwischen den „first-order relations„ und der holistischen Verarbeitung befindet (Reed et al., 2006; siehe Abbildung 1). Für die Gesichterverarbeitung scheinen alle vier Ebenen von Bedeutung zu sein. Bisherige Studien konnten zeigen, dass die ersten beiden Ebenen auch bei der Verarbeitung menschlicher Körperformen eine zentrale Rolle spielen. Belege für die Bedeutung holistischer Verarbeitungsmechanismen und „second-order relations“ für die Körperwahrnehmung fehlen bislang. Demnach teilen sich Gesichter und Körper bis zu einem gewissen Grad einen gemeinsamen Verarbeitungspfad („first-order relations“ und hierarchische Struktur), wobei Unterschiede in späteren Verarbeitungsphasen auftreten können (holistische Verarbeitung und „second-order relations“). Die holistische Verarbeitung wird im experimentellen Kontext meist mit der „composite illusion“ untersucht (siehe Abbildung 3 und Abschnitt 1.1.1). Belege für die Existenz oder das Fehlen eines vergleichbaren Effektes für Körperstimuli könnten weitere Hinweise für Parallelen oder Unterschiede bei der Verarbeitung von Gesichtern und Körpern liefern.

1.7.2 Ziele der Arbeit Das Ziel der ersten Studie war es, mittels fMRT zu neuen Erkenntnissen über die neuronalen Mechanismen zu gelangen, welche der Gesichterverarbeitung zugrunde liegen. Bei der Wahrnehmung menschlicher Gesichter wird ein weit verbreitetes Netzwerk aktiviert. Die gemeinsame Aktivität der Regionen dieses Netzwerks scheint für die erfolgreiche Gesichterverarbeitung notwendig zu sein (Fairhall & Ishai, 2007; Ishai, 2008). Mit Hilfe einer psychophysiologischen Interaktionsanalyse (PPI-Analyse) sollte die funktionelle Verknüpfung der gesichtsselektiven Hirnareale untersucht werden. Auf der Basis vorheriger Untersuchungen (Haxby et al., 2000; Kanwisher et al., 1997; Rossion et al., 2000a) wurde die Hypothese


formuliert, dass die Aktivierung der rechten FFA mit den Aktivierungen der anderen gesichtssensitiven Hirnareale in Verbindung steht. Darüber hinaus sollten auf der Grundlage dieser Befunde die neuronalen Mechanismen, die ein selektives Gesichterverarbeitungsdefizit (Prosopagnosie) auslösen, genauer verstanden werden. Zu diesem Zweck wurden die gesichtsspezifischen Aktivierungsmuster gesunder Kontrollprobanden mit denen von Personen mit einer angeborenen Prosopagnosie verglichen. In einer vorherigen elektrophysiologischen Untersuchung an den gleichen Patienten wurden ein heterogenes Verhaltensmuster sowie heterogene N170-Amplituden und -Latenzen beobachtetet (Minnebusch et al., 2007). Aufgrund dessen wurde in dieser Studie angenommen, dass Personen mit einer Prosopagnosie ebenfalls ein heterogenes Aktivierungsmuster in der FFA und der OFA zeigen würden. In der zweiten Studie sollten die Mechanismen untersucht werden, die an der Wahrnehmung menschlicher Körper und Tiere beteiligt sind. Vorherige Studien deuten darauf hin, dass menschliche Körper, ähnlich wie menschliche Gesichter, konfigural verarbeitet werden. Hinweise darauf, dass Gesichter konfigural verarbeitet werden, stammen hauptsächlich von Untersuchungen des Inversionseffektes (Yin, 1969; 1970). Der Inversionseffekt manifestiert sich sowohl auf der Verhaltensebene als auch auf elektrophysiologischer Ebene. In dieser Studie wurde an gesunden Probanden mittels EEG und Verhaltensmessungen untersucht, ob der Inversionseffekt auch bei der Wahrnehmung menschlicher Körper auftritt und ob es Unterschiede bei der Wahrnehmung menschlicher Körper und Tierkörper gibt. Darüber hinaus wurde untersucht, welchen Einfluss die Darbietung des Kopfes auf die Wahrnehmung menschlicher Körper hat. Die bislang einzigen EEG-Studien, die den Körperinversionseffekt untersucht haben, berichten bei der invertierten Darbietung von Körpern mit Köpfen größere N170-Amplituden sowie eine verzögerte Latenz verglichen mit der aufrechten Darstellung (Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004). Ein Ziel dieser Untersuchung bestand darin, diese Effekte sowohl bei Körpern mit als auch ohne Kopf zu replizieren. Zudem konzentrierten sich die vorherigen Studien ausschließlich auf elektrophysiologische Maße. Ein weiteres Ziel war daher die Kombination elektrophysiologischer und behavioraler Inversionsmaße, welche eine differenziertere Betrachtung zuvor berichteter Effekte erlauben. Wenn es sich bei dem Körperinversionseffekt um ein Phänomen handelt, welches spezifisch für menschliche Körper ist, sollte dieser Effekt bei Tieren nicht auftreten. Alternativ wurde die Hypothese formuliert, dass dieser Effekt bei Tieren auftritt, deren Körperform der menschlichen Körperform ähnelt. Ziel der dritten Studie war eine genauere Charakterisierung des Zusammenhangs zwischen dem Rotationswinkel von Körperstimuli und dem Grad der Störung der konfiguralen Verarbeitung, sowie seines möglichen neuronalen Korrelats. Eine frühere Studie deute auf einen quadratischen Zusammenhang zwischen dem Rotationswinkel eines Gesichts und der Beeinträchtigung der konfiguralen Verarbeitung


hin, was sich sowohl auf der Verhatensebene wie auch in den elektrophysiologischen Korrelaten zeigte (Jacques & Rossion, 2007). Darauf basierend wurde ein quadratischer Zusammenhang zwischen dem Rotationswinkel und der Störung der konfiguralen Verarbeitung angenommen: Amplitude und Latenz der N170/N190 sollten mit zunehmendem Rotationswinkel ansteigen und ab einem Winkel von ungefähr 180 Grad stetig wieder abfallen. Aktuelle Studien zum Inversionseffekt legen eine konfigurale Verarbeitung menschlicher Körperformen nahe (Maurer et al., 2002). Unklar ist, welche der drei bzw. vier Ebenen der konfiguralen Verarbeitung (siehe 1.1) der Körperwahrnehmung zugrunde liegt. In einer weiteren Studie wurde anhand des „composite effects“ untersucht, ob Körper - ähnlich wie Gesichter - holistisch verarbeitet werden. Wenn Körper holistisch verarbeitet werden, dann sollten bei gesunden Probanden bei der Verarbeitung verschobener Körper im Vergleich zur nicht verschobenen Darstellung weniger Fehler und schnellere Reaktionszeiten zu beobachten sein. In einem Übersichtsartikel wurden die Befunde sowohl für als auch gegen die Existenz eines Mechanismus, welcher auf die Verarbeitung menschlicher Körper spezialisiert ist, zusammengefasst und kritisch diskutiert. Die kognitiven Mechanismen, neurokognitiven Modelle und neuronalen Korrelate der Körperverarbeitung sollten mit den Erkenntnissen der Literatur der Gesichterverarbeitung verglichen werden. Zudem sollten in diesem Zusammenhang die neurophysiologischen Prozesse der Körperwahrnehmung und klinische Störungen, die mit Veränderungen der Körperwahrnehmung in Zusammenhang stehen, beschrieben werden. Das Hauptziel der vorliegenden Arbeit war, neue Erkenntnisse über die neurokognitiven Mechanismen der Gesichter- und Körperverarbeitung zu gewinnen, wobei Parallelen und Unterschiede zwischen den Verarbeitungsprozessen im Fokus des Interesses standen.

STUDIEN

UND

ÜBERSICHTSARTIKEL I ZWEI SEITE 28

Studie 1

A bilateral occipitotemporal network mediates face perception Denise A. Minnebusch, Boris Suchan, Odo Köster und Irene Daum (2009a) Behavioural Brain Research, 198, 179-185

STUDIEN

UND


Author's personal copy

Behavioural Brain Research 198 (2009) 179–185

Contents lists available at ScienceDirect

Behavioural Brain Research journal homepage: www.elsevier.com/locate/bbr

Research report

A bilateral occipitotemporal network mediates face perception Denise A. Minnebusch a,∗ , Boris Suchan a , Odo Köster b , Irene Daum a a b

Institute of Cognitive Neuroscience, Department of Neuropsychology, Ruhr-University of Bochum, Universitätsstraße 150, D-44780 Bochum, Germany Institute for Diagnostic and Interventional Radiology and Nuclear Medicine, St. Josef Hospital Bochum, Bochum, Germany

a r t i c l e

i n f o

Article history: Received 10 September 2008 Received in revised form 22 October 2008 Accepted 27 October 2008 Available online 11 November 2008 Keywords: Developmental prosopagnosia FFA Functional connectivity OFA PPI

a b s t r a c t The aim of the present study was to further explore the neuronal mechanisms of face processing in healthy subjects which may help to understand the difficulties experienced by prosopagnosia subjects. A further goal was to compare face specific activation patterns in the right and left occipital face area (OFA) and fusiform face area (FFA) for famous faces, non-famous faces and caricatures of famous faces in four individuals suffering from developmental prosopagnosia (DP) and seven healthy controls, using functional magnetic resonance imaging and psychophysiological interaction analysis (PPI). Control subjects showed higher face related activations in the right compared to the left FFA. Caricatures of faces of famous people and photographs of non-famous faces yielded higher percent signal changes in the OFA and FFA compared to photographs of famous faces. These data support the idea that the OFA and FFA discriminate between familiar and new face representations. The activation patterns of DP subjects were heterogeneous, with none of the patients showing bilateral face related activations in both OFA and FFA. There was no evidence of a left hemispheric activation when the right homologue failed to be activated, supporting the view of a right hemispheric dominance in face perception. PPI analysis indicated a link between activation of the right FFA and the other three tested regions, the left FFA and the right and left OFA. In summary, all four face related brain regions appear to be necessary for successful face processing, and disruption of one component may lead to face recognition deficits. © 2008 Elsevier B.V. All rights reserved.

1. Introduction Recent studies suggest that face and object recognition involve qualitatively different processes which may depend upon distinct brain areas. Patients with brain lesions may show selective impairments of face recognition compared to the recognition of objects, and the opposite pattern has also been described [28,30,42,43]. The double dissociation implies that different brain regions are recruited for the recognition of faces and other types of visual objects [1,12,12]. Functional neuroimaging studies have yielded a region in the fusiform gyrus, the “fusiform face area” (FFA) which is activated more strongly during passive viewing of faces compared to object stimuli [20,23,25]. Bilateral FFA activation correlates with successful face processing, but not with successful object processing [16], with higher correlations in the right hemisphere [21,23,36]. In addition to the FFA, the superior temporal sulcus (STS) and the inferior and mid occipital gyri are also active during face processing [21,23], also with a right hemispheric dominance. The face selective inferior

∗ Corresponding author. Tel.: +49 234 32 23574; fax: +49 234 32 14622. E-mail address: [email protected] (D.A. Minnebusch). 0166-4328/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.bbr.2008.10.041

occipital area termed “occipital face area” (OFA; [21]) is sensitive to physical features of face stimuli, whereas the STS appears to process dynamic aspects of facial information (e.g. expression, gaze). Although there is consensus that the FFA is selectively involved in face perception and identification [16,23,25], its precise functional role remains to be fully explored. There is some evidence that the FFA is linked to view dependent face representations [31]. Haxby et al. [20] have proposed a hierarchical model of face perception, based on a core system and an extended system. The core system involves bilaterally the inferior occipital gyrus (location of the OFA), the lateral fusiform gyrus (location of the FFA) and the STS. The extended system comes into play if additional information is extracted from faces (e.g. facial expression, lip-reading, or name and biographical information), whereas facial expressions are also processed by the STS. The authors suggested that both the FFA and the STS receive input from the OFA. This view was supported by a recent functional magnetic resonance imaging (fMRI) study [11], although there is as yet no evidence for interhemispheric connectivity between the face areas. Additionally, training of configural face processing enhances the functional connectivity between the right FFA and OFA [9]. Rossion et al. [35] suggest that the right OFA is reciprocally connected to the right FFA, with both regions being necessary for intact face perception. In addition, Rossion et al. [34,35] proposed a direct pathway from early visual areas to the

STUDIEN

UND



180

D.A. Minnebusch et al. / Behavioural Brain Research 198 (2009) 179–185

FFA. Connectivity between the OFA and FFA and the nature of the interaction between the right and left FFA and OFA remain to be fully determined. Face perception seems to activate a distributed cortical network in the human brain and the integrated activity of this network appears to be required for successful face perception [11,22]. Psychophysiological interaction analysis (PPI) allows the study of the functional coupling of face selective brain areas which form a highly specialised network. Based on previous findings [20,23,36], the current study aimed to test the hypothesis that activation of the right FFA is strongly associated with the activity of other face sensitive brain regions. Former studies using similar approaches (e.g., see [22]) focused mainly on the posterior–anterior coupling of face related areas using pre-defined parts of a face selective network. DeGutis et al. [9] on the other hand focused on right lateralized functional connectivity of face selective areas. The present PPI analysis aimed at the detection of the functional network of bilateral posterior face selective areas, with the right FFA as pre-defined region of interest. Studies of patients suffering from acquired prosopagnosia (AP) indicated that the functional integrity of the right fusiform gyrus might be necessary for face processing [3,45]. Furthermore, Steeves et al. [43] reported intact fusiform gyrus function in an AP patient, while there was evidence of bilateral OFA damage, suggesting that the functional integrity of a network involving both the FFA and the OFA may be critical for face processing. This implies that either abnormal OFA or FFA activation patterns or both may underlie the deficits in prosopagnosics. Another AP subject has lesions in the right OFA and left FFA but intact left OFA and right FFA [35,42], indicating that bilateral activation in the face network seems to be necessary for successful face perception [22]. In addition to AP, lifelong face processing impairments have been observed in the absence of perinatal or early brain injury, a condition termed “developmental prosopagnosia” (DP; [2,5]). Only few DP cases have been investigated with measures of neural activity so far and knowledge about the underlying mechanism is sparse [4,26]. Behavioural [27] and electrophysiological studies [18,29] have indicated that DP reflects a heterogeneous impairment. A recent study [6] combined fMRI and electrophysiological methods to provide a comprehensive description of a DP subject. In this case, face-selectivity was absent in both fMRI and electrophysiological assessments. However, a previous DP subject showed normal face specific FFA activations but abnormal face related electrophysiological activation pattern [30]. Both biological markers thus appear to be related to different perceptual mechanisms [6]. A face specific component in EEG, the N170 is activated while physiognomic stimuli occur in the visual field, whereas face specific fMRI activation might be associated with configural face processing. The available evidence for FFA and OFA activations in prosopagnosia is inconclusive. Hadjikhani and de Gelder [17] did not find face related FFA responses in two DP and one AP subject, whereas Hasson et al. [19] reported face related activation patterns in the ventral occipito-temporal cortex which were similar to controls. Furthermore, Bentin et al. [6] did not find face related activation in the fusiform gyrus or posterior temporal lobes (including the FFA) of a DP subject which might suggest global processing deficits. In a further study, four DP subjects showed fMRI activation patterns in the fusiform gyrus in response to faces, buildings and other objects [1]. Comparable results were reported in a recent single study [10]. The dissociation between normal FFA activation and behavioural face processing impairment implies that FFA activation is not sufficient for successful face processing [18,27]. Whether processing of famous and non-famous faces in healthy subjects is critically linked to the fusiform gyrus is another issue which remains to be fully explored [10,15,32,37,38]. Two recent studies demonstrated differential activations in the right OFA and

FFA for familiar and unfamiliar faces [37,38]. Additionally, a face specific component in EEG, the N170, is larger for non-famous faces compared to famous faces and caricatures of famous faces in both healthy subjects and some but not all DP subjects [29]. Behavioural data suggest that healthy subjects tend to find it easier to recognise famous people on the basis of photographs compared to caricatures, while this effect is not observed in prosopagnosia [8,29]. In addition, three of the DP subjects, who participated in the present study showed a trend towards better recognition of caricatures compared to photographs of famous people (see Table 1). In caricatures, facial characteristics are overstated and DP subjects might benefit from the exaggerated display of facial features. To further address the issue of potential processing differences across different types of face stimuli, the present study assessed OFA and FFA activation patterns for famous faces, non-famous faces and caricatures of famous faces in both healthy subjects and four individuals suffering from DP. Given the inconsistent findings for activation patterns in prosopagnosics compared to controls [6,17] and the considerable heterogeneity of impairment patterns in prosopagnosia (e.g. [29]), the study of individual prosopagnosia subjects relative to matched controls appears to be a promising approach to gain further insight into the mechanisms of breakdown of face processing. A second goal of the fMRI study was to elucidate interactions between specific face processing areas in healthy subjects and their specific contributions to the processing of different representations of faces such as photographs and caricatures of famous people. 2. Method 2.1. Subjects Four developmental prosopagnosics (two female; mean age 37.3 years, SD = 2.5) and seven healthy control subjects (three female; mean age 42.7 years, SD = 9.8) participated in this study. The prosopagnosic subjects (ET, LT, NN and TP) have been described in detail in a recent report [29]. All subjects were right-handed and received reimbursement for their expenses for participation. All participants had normal or corrected-to-normal vision. Structural MRI (T1) did not reveal any overt brain abnormalities in the prosopagnosic group. The study procedure was approved by the Ethics Committee of the Medical Faculty of the Ruhr-University, Bochum, Germany and was performed in accordance with ethical standards laid down in the Declaration of Helsinki [44]. Written informed consent was obtained from all subjects. 2.2. Behavioural assessment For detailed description of the behavioural assessment and the specificity of the impairment pattern see Minnebusch et al. [29]. Individual test scores and control data for object and face processing performance are presented in Table 1. To sum up, none of the prosopagnosics were impaired in tests of low level vision and basic level object recognition (Birmingham Object Recognition Battery [33]; Benton Battery [7]; line drawings from the pictures of Snodgrass and Vanderwart [41]). However, all prosopagnosics were impaired in memory for faces (Recognition Memory Test, subtest Faces; [46], the learning of new faces (Bochum Test of Face Processing Skills; [29]) and the recognition of famous faces. The behavioural data thus demonstrated significant face processing deficits in all prosopagnosic subjects. 2.3. Stimuli for fMRI study Grey-scaled pictures (140 Pixel × 200 Pixel) belonging to different categories (houses, non-famous faces, famous faces, famous caricatures, musical instruments, tools, butterflies, fruits and vegetables, cars, bicycles, sunglasses, guns, horses, faces of horses and horses in movement) were presented in a block-design fashion. Each stimulus class consisted of 40 photographs, with the exception of “houses” which contained 32 pictures (for a more detailed description, see [29]). Photographs of non-famous faces and famous faces were taken from front views. They had a neutral expression and only internal features were visible (see Figs. 1 and 3). Photographs for the non-famous face category were taken from a set of photographs of 20 males and 20 females, which was developed in our department. Pictures for the famous faces and non-famous faces categories were selected from the Internet and consisted of photographs of 27 male and 13 female celebrities. The caricatures consisted of greyscaled illustrations (see Fig. 1). The face of each famous person was once used as a stimulus in the realistic photograph and once in the caricature category. In pilot

STUDIEN

UND



181

D.A. Minnebusch et al. / Behavioural Brain Research 198 (2009) 179–185 Table 1 Demographical data and performance of the prosopagnosic subjects in object (a) and face processing tasks (b). (a) ET

LT

NN

TP

Age Sex

41 Male

36 Female

36 Male

36 Female

BORB Length matching Size matching Orientation matching Gap matching

Normal Normal Normal Normal




Benton Right-left orientation Judgment of line orientation Visual form discrimination

Normal Normal Normal




Snodgrass

Normal

Normal

Normal

Normal

Rey figure Copy Recall

Normal Normal

Normal Normal

Normal Affected

Normal Normal

(b) ET

LT

NN

TP

Controls

100 92.9 93.3 93.3 93.3

98.8 (±2.7) 91.6 (±5.6) 94.8 (±5.8) 96.5 (±3.7) 94.3 (±5.3)

TAB Identity discrim. Affect discrim. Affect naming Affect identification Affect matching

93.3* 91.6 94.8 86.6* 94.3

91.6* 78.6* 86.7* 86.7* 86.7*

91.6* 85.7* 66.7* 93.3 46.7*

BFRT RMT-F

45 38*

44 35*

40* 33*

42* 34*

Famous faces 1 Famous faces 2 Famous caricatures

35.7* 44.8* 24.1*

10.5* 23.7* 28.9*

13.6* 29.0* 34.21*

10* 13.9* 25.0*

Bochumer test of face processing d’ RT for hits (ms)

0.6* 117

0.09* 380*

1.9 436*

1.3 430*

NR 43-46 43.9 (± 3.9a ) 44.8 (±3.3b ) 76 (±14) 84 (±12) 63 (±15) 2.1 (±0.7) 187 (±99)

NR, normal range; d’, measure of discrimination; RT, reaction time; a,b age related controls data, a 35–39 years, b 40–44 years. * Performance below average (z < 1.65 relative to control group).

studies involving young healthy controls, 84% (SD = 11.8) of the photographs and 63% (SD = 15.0) of the caricatures were correctly named. All prosopagnosics were impaired in recognising famous people on the basis of both realistic photographs (scores: 13.9–44.5%) and caricatures (scores: 24.1–34.2%; see [29]).

The subjects were specifically instructed to carefully study the photographs, while working on the secondary task.

2.4. Procedure

520 scans consisting of 25 slices with TR = 2000 ms (TE = 40 ms, FOV = 192 mm, flip angle = 90◦ ) and a 1 mm gap were acquired for each subject on a Siemens (www.siemens.de) Symphony 1.5 T scanner. The voxel size was 3 mm × 3 mm × 3.6 mm. A high resolution T1 image (MP-RAGE sequence) of 112 slices was additionally acquired for anatomical labeling (TR = 1990 ms, TE = 4.18 ms, FOV = 250 mm, flip angle = 15◦ ). The voxel size was 1.4 mm × 1 mm × 1.5 mm. Functional images were analysed using SPM 5 (http://www.fil.ion.ucl.ac.uk/spm). The first five images were discarded to allow for T1 equilibration. The analysis started with the slice timing procedure followed by the realignment of all images to the first of a series. Images were normalised to the stereotactic space of the Montreal Neurological Institute (MNI) brain provided by SPM 5, and smoothed with a Gaussian kernel of 8 mm. A GLM was applied to the data separating them according to the stimulus category. Data were high-pass filtered to eliminate low-frequency components (a cut-off value of 128 s was used).

Stimuli were projected onto a screen in front of the MR scanner and the subjects viewed the screen via a mirror mounted on the head-coil. Pictures belonging to different categories including photographs of non-famous faces, famous faces and famous caricatures were presented on a white background, with a black cross placed in the centre. Each stimulus was presented for 500 ms with an inter stimulus interval (ISI) of 1500 ms (Fig. 1). During the ISI a black cross (baseline epochs) appeared in the centre of the monitor. Pictures were presented in a block-design fashion. Each block contained 20 pictures and lasted 40 s. There were two runs of each block, presented in a random order. Subjects were instructed to press a key when a small circle placed in the bottom right quadrant of one picture was followed by a large circle in the next picture. This task was used to maintain attention to the stimuli without explicitly addressing perceptual processing of the stimuli being viewed.

2.5. Scanning procedure and analysis

Fig. 1. Experimental design.

STUDIEN

UND



182

D.A. Minnebusch et al. / Behavioural Brain Research 198 (2009) 179–185 Table 2 MNI (Montral Neurological Institute) coordinates of the face related activation in the inferior occipital and fusiform gyrus of control subjects. R/L Controls AM

R L R L R L R L R L R L R L R L

BM KT RR RSB

Fig. 2. OFA and FFA in healthy controls.

SA TSB

Region of interest (ROI) were localised individually. In agreement with previous studies [23–25,39] the contrast [(famous faces + non-famous faces + famous caricatures) − 3 (houses)] was used to determine the FFA and OFA in all healthy controls and DP subjects, using a cluster of at least eight contiguous voxels (p < 0.001, uncorrected). The statistical threshold had to be lowered to p < 0.01 (uncorrected) to determine the FFA and OFA in some DP subjects. Percentage signal changes were extracted for each subject using the marsbar software package (http://marsbar.sourceforge.net/). The percentages of signal changes within these brain regions were extracted for the experimental conditions “photographs of famous faces”, “photographs of non-famous faces” and “caricatures of famous faces”. The same approach was used for the analysis of the DP data. If a FFA or OFA region could be identified, percentages of signal changes were determined for these regions for the three stimulus types (non-famous faces, famous faces and caricatures). The data of the DP individuals were converted into z-scores relative to control subjects (cut-off ± 1.65). 2.6. Psychophysiological interaction (PPI) One primary aim of the present study focused on the assessment of interactions among the network responsible for face processing including fusiform and occipital face areas. Based on results from the literature [21,23,36], we hypothesised a critical involvement of the right FFA in face processing as well as a possible modulatory contribution of activity in other brain regions. The aim of this analysis was to identify the areas which show a strong covariation with the right FFA during face processing and a weak covariation during the processing of houses. In other words, we aimed to identify brain activations which could be explained by an interaction between experimental factors (faces vs. houses) and the right FFA activation. We therefore performed an exploratory PPI analysis [13] implemented in SPM5 for the data of the control subjects. As the DP subjects showed very heterogeneous activation patterns in the FFA and the OFA (see Table 3), PPI might provide some insights into the face processing network in healthy controls which in turn should offer possible explanations for the patterns seen in DP subjects. A sphere of 8 mm diameter was placed over the individual right FFA as a region of interest and the contrast of non-famous faces vs. houses was used in PPI analysis.

3. Results 3.1. Localisation of FFA and OFA For all control and DP subjects, we initially localised face related activations by identifying voxels which were selectively activated by faces (famous faces, non-famous faces and caricatures) compared to houses [39]. Consistent with previous studies [20,23], face related activations in healthy controls emerged in a region within the fusiform gyrus (Fig. 2), corresponding to the FFA and in the lateral inferior occipital gyrus, corresponding to the previously described OFA [14]. The individual coordinates for face related regions (FFA and OFA) in both hemispheres are presented in Table 2. Localisation of FFA and OFA in prosopagnosics yielded a heterogeneous pattern (Table 3). ET showed a right and left FFA and a left OFA activation, whereas LT did not show face related responses in the FFA and OFA in both hemispheres. In NN, faces activated the right FFA as well as the right OFA, whereas TP showed face related activation only in the right OFA.

Mean (SD)

FFA

OFA

x

y

z

x

y

z

44 −36 34 −38 44 −42 40 −34 40 −32 42 −38 38 −40 40 (4) −37 (3)

−52 −54 −44 −48 −62 −62 −48 −42 −60 −46 −54 −52 −52 −56 −53 (6) −51 (7)

−22 −20 −24 −26 −16 −16 −20 −20 −24 −26 −16 −20 −18 −20 −20 (4) −21 (4)

38 −44 36 −38 42 −32 36 −36 46 −42 48 −42 48 −30 42 (5) −38 (5)

−72 −82 −75 −84 −78 −84 −78 −68 −78 −70 −70 −70 −82 −90 −76 (4) −78 (9)

−12 −10 −12 −12 −8 −6 −16 −10 −10 −20 −14 −8 −6 −8 −10 (3) −11 (5)

L, left hemisphere; R, right hemisphere.

3.2. Activation in the FFA and OFA during processing of famous faces, non-famous faces and caricatures of famous faces in control subjects Percent signal changes were further analysed using repeated measures ANOVA, separately for OFA and FFA, with factors hemisphere (right vs. left) and faces (famous faces, non-famous faces and caricatures). Results of the signal changes in the FFA yielded a significant main hemisphere effect (F(1,7) = 13.5; p = 0.01) with higher percent signal changes in the right (M = 0.12, SD = 0.06) compared to the left hemisphere (M = 0.06, SD = 0.05) and a trend toward a significant main face effect (F(1,7) = 3.4; p = 0.09) with higher percent signal changes for caricatures and non-famous faces compared to famous faces (see Fig. 3). Analysis of the percent signal changes in the OFA yielded a significant main faces effect (F(1,7) = 5.8; p < 0.03) with higher percent signal changes for caricatures of famous faces (M = 0.12, SD = 0.06) and non-famous faces (M = 0.10, SD = 0.06) compared to photographs of famous faces (M = 0.06, SD = 0.04). None of the other comparisons reached significance. 3.3. Percent signal changes in prosopagnosia Percent signal changes in the FFA and OFA for photographs and caricatures of famous faces and photographs of non-famous faces are illustrated in Fig. 3. The pattern of FFA and OFA signal changes in DP subjects is heterogeneous. ET showed a higher signal change in the right FFA for famous faces compared to caricatures of famous Table 3 MNI (Montral Neurological Institute) coordinates of the face related activation in the inferior occipital and fusiform gyrus of DPs. R/L DPs

FFA

OFA

x

y

z

x

y

z

ET

R L

38 −46

−66 −66

−6 −6

– −48

– −74

– 0

LT

R L

– –

– –

– –

– –

– –

– –

NN

R L

44 –

−50 –

−18 –

48 –

−72 –

−10 –

TP

R L

– –

– –

– –

50 –

−66 –

−18 –


STUDIEN

UND



183


Fig. 3. Percent signal changes in the FFA and OFA for each DP subject and average activation profiles of control subjects for non-famous faces, famous caricatures and famous faces.

faces and non-famous faces. NN showed the opposite pattern, with higher signal changes in the right FFA and OFA during the processing of caricatures and non-famous faces and a reduction during the processing of famous faces (see Fig. 3). Activation patterns in the left FFA were comparable for ET and controls. In the right OFA, TP showed higher signal changes for photographs of famous and nonfamous faces compared to caricatures (see Fig. 3). Because of the large variability and the small sample size, the data of the prosopagnosics were analysed in terms of z-scores relative to the control group. For ET, the activation levels for famous faces (z = −1.7) and non-famous faces (z = −2.8) were reduced in the left OFA. TP showed higher activations in the right OFA for non-famous faces compared to controls (z = 2.1). The activation in the right FFA for photographs of famous faces was enhanced in ET (z = 3.1). NN showed reduced activation for famous faces in the right FFA (z = −3.6). None of the other comparisons reached significance (see Fig. 3). 3.4. Psychophysiological interaction analysis in the control group The results of the PPI analysis are listed in Table 4. Activations in the left FFA and the left OFA as well as in the right OFA yielded significant interactions with right FFA activation. An overlay of the localiser activation and results from the PPI analysis (see Tables 1 and 3) indicate overlapping locations in the left FFA and OFA. Additional activations were observed bilaterally in the cuneus

Table 4 Results of the PPI analysis. p < 0.001 (uncorrected at voxel level; p < 0.01 corrected for family wise errors (FWE) at cluster level). Cluster size

z-value

R/L

54

4.39 4.01 4.28 4.22 4.14 4.07 3.99 3.96 3.94 3.90 3.79 3.19

R R R L L L R R L L L R

22 24 40 15 9 30 20 17 9

x 30 38 62 −20 −22 −34 22 32 −38 −18 −14 36

y

z

−84 82 48 −4 − −84 −74 −92 −54 −80 −106 −74

30 30 10 0 8 −14 −12 14 8 4 6 −12

Cuneus, BA 19 Superior occipital gyrus, BA 19 Superior temporal gyrus, BA 22 Lateral globus pallidus Putamen Left occipital face area Lingual gyrus, BA 18 Middle occipital gyrus, BA 19 Left fusiform face area Lingual gyrus, BA 18 Cuneus, BA 18 Right occipital face area


and lingual gyrus. The right superior temporal and the middle occipital gyrus were also activated, as were the left globus pallidus and the putamen. 4. Discussion The aims of the present study were to further clarify the functional neuroanatomy of face processing in healthy humans and to elucidate dissociations between impaired face processing abilities

STUDIEN

UND



184


and normal activation in face related brain areas (FFA and OFA) in DP subjects. Furthermore, we investigated potential differential processing patterns of caricatures, famous and non-famous faces in prosopagnosia. Results of an exploratory PPI analysis yielded evidence for a close relationship between activity in the right FFA and the left FFA and the OFA bilaterally in healthy subjects. In other words, activation of the right FFA might selectively strengthen the coupling with those areas implicated in face processing without using pre-defined additional regions of interests. Anatomical connectivity between FFA and OFA is still under debate. While some authors assume that the right FFA receives input from the right OFA (see [20]), others have suggested that the right OFA also receives feedback input from the right FFA [34,35]. Fairhall and Ishai [11] showed that the OFA influences the FFA in both hemispheres. Interhemispheric connectivity between the face related areas was not explored. The PPI analysis based on the current data suggest that activation of the right FFA is linked to both the left and right OFA and also to the left FFA. Face specific areas in both hemispheres thus appear to be implicated in face processing, although it remains to be determined whether these regions are necessary for successful face recognition. Analysis using techniques like Dynamic Causal Modelling might provide further insights into the connectivity patterns in question. In addition, Diffusion Tensor Imaging might help to identify anatomical connectivity between these four face associated extrastriate regions. In the absence of interhemispheric connectivity, parallel processing of face selective information in both hemispheres might serve as an alternative hypothesis. Additional covariations were found in visual related areas like the cuneus, lingual gyrus and middle occipital gyrus which presumably reflect additional face related processing of the stimuli. The functional contribution of these areas remains to be determined. To our knowledge, this is the first study which compares activation patterns in the FFA and OFA linked to the processing of photographs of famous and non-famous faces and caricatures of famous people. In previous studies [8,29], subjects consistently showed better performance for photographs compared to caricatures. In other words, it appears to be easier to identify familiar faces on the basis of photographs than on the basis of caricatures. Caricatures comprise the exaggerated display of characteristical features, and the overstated representation of the facial characteristics might induce distinct activation profiles in face related areas. Previous findings concerning a differential FFA involvement in the processing of familiar and unfamiliar faces are inconsistent [10,15,32,37,38]. The present data imply higher FFA and OFA activation patterns for caricatures and non-famous faces than for famous faces (Fig. 3). The OFA is presumably involved in the basic level analysis of facial features [20] which according to our data is significantly affected by the type of face representation. The present results are consistent with a previous PET study [37] which suggested that the OFA and FFA discriminate novel from visually familiar faces. It is possible that face related areas discriminate familiar face representations (here: photographs of famous faces) from novel representations (here: photographs of non-famous faces and caricatures of famous faces). Caricatures might be more difficult to process than photographs of famous faces because of our comprehensive everyday expertise with realistic face representations. In addition, forming a representation of a new face involves more effort than perception of a familiar face, for which such a representation is already available. Both assumptions are consistent with previous findings which indicate better recognition performance for realistic representations of famous faces compared to caricatures and an enhanced parieto-occipital N170 for non-famous faces compared to famous faces [29]. Additionally, right FFA acti-

vation was higher than left FFA activation, whereas face related activation patterns were comparable in the left and right OFA. Previous studies have frequently reported right hemisphere dominance for both areas [21,23]. The current findings in healthy subjects are of considerable importance for understanding the heterogeneous FFA and OFA activation patterns in prosopagnosia. None of the DP subjects showed bilateral face related FFA and OFA activations, whereas this was the case in all control subjects. Furthermore, one DP subject (LT) failed to show any face related activation either in the FFA or in the OFA, and there was no systematic activation pattern in these regions in the remaining three prosopagnosic subjects. There seems to be, however, a right hemisphere bias since a left hemispheric region was never activated in the absence of activation in the right homologue (see Table 3). This result offers some support to the assumption that the right hemisphere plays a dominant role in face processing [11,20], although PPI analysis clearly indicates that the functional integrity of all four regions is necessary for intact face processing. ET showed face related activations in the right FFA but not in the right OFA which supports the view that the FFA can be activated independently of the OFA [34]. Interestingly, the two prosopagnosic subjects with the best performances in the famous face recognition task (ET and NN) showed activations of both FFA and OFA in at least one hemisphere, whereas the DP subjects with lower performances showed only OFA activation (TP) or no activation at all (LT). To further elucidate the selectivity of the processing deficit in the prosopagnosics, the FFA and OFA activation profiles for the different face stimuli were determined for each individual DP subject and compared to the control data. ET showed reduced activations for famous and non-famous faces in the left OFA, but normal activation patterns in the left FFA. He did not show face related activations in the right OFA, but enhanced activations for famous faces in the right FFA. ET’s face processing difficulties might thus result from inadequate basic analysis of facial features in the OFA. The hierarchical face processing model suggests that the FFA receives input from the OFA [11,20]. Rossion [34] has pointed out that there might be a direct pathway from the early visual cortex to the FFA. However, input from the OFA as well as an OFA and FFA interaction are necessary for face identification [34,40,43]. For NN, photographs of famous faces were associated with reduced activity in the right OFA and the right FFA, whereas nonfamous faces and caricatures evoked normal activation patterns in the right OFA and FFA. Rossion [34] suggested that the right OFA performs finer-grained perceptual analyses, which might be guided by the activation of a face template in the right FFA, and normal face processing abilities might thus depend upon an interaction between both areas. NN’s difficulties in recognising familiar people might thus be linked to reduced activations in the right OFA and FFA. The observed normal activations for non-famous faces and famous caricatures suggest that reliable right hemisphere activations are not sufficient for normal face processing. In TP, face related activation occurred only in the right OFA, which was enhanced for non-famous faces. Since face perception and identification in healthy subjects are associated with right FFA activation [16,23,25], the absence of this effect might underlie the face processing deficit in TP. Photographs and caricatures of famous people evoked normal activation pattern in the right OFA, an effect which is apparently not sufficient for normal face identification since according to PPI analysis, the functional integrity of right and left FFA and OFA is required for intact face processing skills. None of the DP subjects showed abnormal brain activation patterns for caricatures of famous people, indicating that prosopagnosics might benefit from illustrations which overstate characteristical facial features.

STUDIEN

UND




In sum, previous investigations of face processing have mainly focused on face related activation patterns in the right FFA and the right OFA. The current PPI analysis indicates that bilateral OFA activations and left FFA activation show a significant covariation with the right FFA. The functional integrity of this network appears to be necessary for successful face processing. Disruption of function in at least one of these areas and/or changes in anatomical connectivity might underlie the face processing impairments in prosopagnosia. This idea is at least partly supported by findings of Rossion et al. [38] and Steeves et al. [43], which emphasised the importance of the functional integrity of both the OFA and the FFA for normal face processing. Our data are also consistent with the face perception model proposed by Haxby et al. [20] suggesting that a distributed and bilateral neural system which includes both FFA and OFA mediates the visual analysis of faces. References [1] Avidan G, Hasson U, Malach R, Behrmann M. Detailed exploration of facerelated processing in congenital prosopagnosia: 2. Functional neuroimaging findings. J Cogn Neurosci 2005;17:1150–67. [2] Barton JJ, Cherkasova MV, Press DZ, Intriligator JM, O’Connor M. Developmental prosopagnosia: a study of three patients. Brain Cogn 2003;51:12–30. [3] Barton JJ, Press DZ, Keenan JP, O’Connor M. Lesions of the fusiform face area impair perception of facial configuration in prosopagnosia. Neurology 2002;58:71–8. [4] Behrmann M, Avidan G. Congenital prosopagnosia: face-blind from birth. Trends Cogn Sci 2005;9:180–7. [5] Behrmann M, Avidan G, Marotta JJ, Kimchi R. Detailed exploration of facerelated processing in congenital prosopagnosia: 1. Behavioral findings. J Cogn Neurosci 2005;17:1130–49. [6] Bentin S, Degutis JM, D’Esposito M, Robertson LC. Too many trees to see the forest: performance, event-related potential, and functional magnetic resonance imaging manifestations of integrative congenital prosopagnosia. J Cogn Neurosci 2007;19:132–46. [7] Benton AL, Sivan AB, Hamsher K, Varney NR, Spreen O. Contribution to neuropsychological assessment. New York: Oxford University Press; 1983. [8] De Gelder B, Rouw R. Configural face processes in acquired and developmental prosopagnosia: evidence for two separate face systems? Neuroreport 2000;11:3145–50. [9] Degutis JM, Bentin S, Robertson LC, D’Esposito M. Functional plasticity in ventral temporal cortex following cognitive rehabilitation of a congenital prosopagnosic. J Cogn Neurosci 2007;19:1790–802. [10] Eger E, Schweinberger SR, Dolan RJ, Henson RN. Familiarity enhances invariance of face representations in human ventral visual cortex: fMRI evidence. Neuroimage 2005;26:1128–39. [11] Fairhall SL, Ishai A. Effective connectivity within the distributed cortical network for face perception. Cereb Cortex 2007;17:2400–6. [12] Farah MJ, Levinson KL, Klein KL. Face perception and within-category discrimination in prosopagnosia. Neuropsychologia 1995;33:661–74. [13] Friston KJ, Buechel C, Fink GR, Morris J, Rolls E, Dolan RJ. Psychophysiological and modulatory interactions in neuroimaging. Neuroimage 1997;6:218–29. [14] Gauthier I, Tarr MJ, Moylan J, Skudlarski P, Gore JC, Anderson AW. The fusiform “face area” is part of a network that processes faces at the individual level. J Cogn Neurosci 2000;12:495–504. [15] Gorno-Tempini ML, Price CJ. Identification of famous faces and buildings: a functional neuroimaging study of semantically unique items. Brain 2001;124:2087–97. [16] Grill-Spector K, Knouf N, Kanwisher N. The fusiform face area subserves face perception, not generic within-category identification. Nat Neurosci 2004;7:555–62. [17] Hadjikhani N, De Gelder B. Neural basis of prosopagnosia: an fMRI study. Hum Brain Mapp 2002;16:176–82. [18] Harris AM, Duchaine BC, Nakayama K. Normal and abnormal face selectivity of the M170 response in developmental prosopagnosics. Neuropsychologia 2005;43:2125–36. [19] Hasson U, Avidan G, Deouell LY, Bentin S, Malach R. Face-selective activation in a congenital prosopagnosic subject. J Cogn Neurosci 2003;15:419–31.

185

[20] Haxby JV, Hoffman EA, Gobbini MI. The distributed human neural system for face perception. Trends Cogn Sci 2000;4:223–33. [21] Haxby JV, Ungerleider LG, Clark VP, Schouten JL, Hoffman EA, Martin A. The effect of face inversion on activity in human neural systems for face and object perception. Neuron 1999;22:189–99. [22] Ishai A. Let’s face it: it’s a cortical network. Neuroimage 2008;40:415–9. [23] Kanwisher N, McDermott J, Chun MM. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J Neurosci 1997;17:4302–11. [24] Kanwisher N, Stanley D, Harris A. The fusiform face area is selective for faces not animals. Neuroreport 1999;10:183–7. [25] Kanwisher N, Yovel G. The fusiform face area: a cortical region specialized for the perception of faces. Philos Trans R Soc Lond B Biol Sci 2006;361:2109–28. [26] Kress T, Daum I. Developmental prosopagnosia: a review. Behav Neurol 2003;14:109–21. [27] Le Grand R, Cooper PA, Mondloch CJ, Lewis TL, Sagiv N, De Gelder B, Maurer D. What aspects of face processing are impaired in developmental prosopagnosia? Brain Cogn 2006;61:139–58. [28] McMullen PA, Fisk JD, Phillips SJ, Maloney WJ. Apperceptive agnosia and face recognition. Neurocase 2000;6:403–14. [29] Minnebusch DA, Suchan B, Ramon M, Daum I. Event-related potentials reflect heterogeneity of developmental prosopagnosia. Eur J Neurosci 2007;25:2234–47. [30] Moscovith M, Winocour G, Behrmann M. What is special about face perception? Nineteen experiments on a person with visual object agnosia and dyslexia but normal face recognition. J Cogn Neurosci 1997;9:555–604. [31] Pourtois G, Schwartz S, Seghier ML, Lazeyras F, Vuilleumier P. Portraits or people? Distinct representations of face identity in the human visual cortex. J Cogn Neurosci 2005;17:1043–57. [32] Pourtois G, Schwartz S, Seghier ML, Lazeyras F, Vuilleumier P. Viewindependent coding of face identity in frontal and temporal cortices is modulated by familiarity: an event-related fMRI study. Neuroimage 2005;24:1214–24. [33] Riddoch MJ, Humphreys GW. Birmingham object recognition battery. UK: LEA: Hove; 1993. [34] Rossion B. Constraining the cortical face network by neuroimaging studies of acquired prosopagnosia. Neuroimage 2008;40:423–6. [35] Rossion B, Caldara R, Seghier M, Schuller AM, Lazeyras F, Mayer E. A network of occipito-temporal face-sensitive areas besides the right middle fusiform gyrus is necessary for normal face processing. Brain 2003;126:2381–95. [36] Rossion B, Dricot L, Devolder A, Bodart JM, Crommelinck M, De Gelder B, Zoontjes R. Hemispheric asymmetries for whole-based and part-based face processing in the human fusiform gyrus. J Cogn Neurosci 2000;12: 793–802. [37] Rossion B, Schiltz C, Crommelinck M. The functionally defined right occipital and fusiform “face areas” discriminate novel from visually familiar faces. Neuroimage 2003;19:877–83. [38] Rossion B, Schiltz C, Robaye L, Pirenne D, Crommelinck M. How does the brain discriminate familiar and unfamiliar faces?: a PET study of face categorical perception. J Cogn Neurosci 2001;13:1019–34. [39] Schiltz C, Rossion B. Faces are represented holistically in the human occipitotemporal cortex. Neuroimage 2006;32:1385–94. [40] Schiltz C, Sorger B, Caldara R, Ahmed F, Mayer E, Goebel R, Rossion B. Impaired face discrimination in acquired prosopagnosia is associated with abnormal response to individual faces in the right middle fusiform gyrus. Cereb Cortex 2006;16:574–86. [41] Snodgrass JG, Vanderwart M. A standardised set of 260 pictures: norms for name agreement, familiarity, and visual complexity. J Exp Psychol Gen 1980;6:174–215. [42] Sorger B, Goebel R, Schiltz C, Rossion B. Understanding the functional neuroanatomy of acquired prosopagnosia. Neuroimage 2007;35:836–52. [43] Steeves JK, Culham JC, Duchaine BC, Pratesi CC, Valyear KF, Schindler I, Humphrey GK, Milner AD, Goodale MA. The fusiform face area is not sufficient for face recognition: evidence from a patient with dense prosopagnosia and no occipital face area. Neuropsychologia 2006;44:594–609. [44] Varga AC. Declaration of Helsinki (Adopted by the 18th World Medical Assembly in Helsinki, Finland, and revised by the 29th World Medical Assembly in Tokyo, 1975). In: The main issue in bioethics. Revised ed. New York: Paulist Press; 1984. [45] Wada Y, Yamamoto T. Selective impairment of facial recognition due to a haematoma restricted to the right fusiform and lateral occipital region. J Neurol Neurosurg Psychiatry 2001;71:254–7. [46] Warrington EK. Recognition memory test. Windsor, UK: NFER-Nelson; 1984.

STUDIEN

UND


Studie 2

loosing your head: behavioral and electrophysiological effects of body inversion Denise A. Minnebusch, Boris Suchan und Irene Daum (2008) Journal of Cognitive Neuroscience, in Druck

STUDIEN

UND


Losing your Head: Behavioral and Electrophysiological Effects of Body Inversion Denise A. Minnebusch, Boris Suchan, and Irene Daum

Abstract & The present study aimed to further explore the mechanisms underlying the perception of human body shapes. Behavioral and electrophysiological inversion effects were studied for human bodies with and without heads and for animal bodies (cats, dogs, and birds). Recognition of human bodies (with heads) was adversely affected by stimulus inversion, and the N170 had longer latencies and higher amplitudes for inverted

INTRODUCTION Recent studies suggest that human faces and human body forms are unique stimulus classes, which are processed in specialized cortical areas (Peelen & Downing, 2007; Kanwisher & Yovel, 2006). Face processing has been linked to the ‘‘fusiform face area’’ (FFA) in the fusiform gyrus (Kanwisher & Yovel, 2006; Haxby, Hoffman, & Gobbini, 2000; Kanwisher, McDermott, & Chun, 1997) and the occipital face area (OFA; Gauthier et al., 2000; Haxby et al., 1999; Kanwisher et al., 1997). Perception of human body forms has also been associated with specific cortical processing units (Peelen & Downing, 2005, 2007; Downing, Chan, Peelen, Dodds, & Kanwisher, 2006; Schwarzlose, Baker, & Kanwisher, 2005; Downing, Jiang, Shuman, & Kanwisher, 2001). The extrastriate body area (EBA) is more strongly activated during perception of human bodies and body parts compared to objects or faces (Urgesi, Berlucchi, & Aglioti, 2004; Downing et al., 2001). The fusiform body area (FBA) in the posterior fusiform gyrus (Peelen, Wiggett, & Downing, 2006; Peelen & Downing, 2005; Schwarzlose et al., 2005) responds to the whole body rather than body parts (Taylor, Wiggett, & Downing, 2007). The EBA seems to be involved in the basic analysis of body features (similar to the OFA in face processing), whereas the FBA (similar to the FFA) might be implicated in processing the configuration of body parts as a whole (Taylor et al., 2007). Faces and human body forms (stimuli presented without heads/ faces) appear to be processed in adjacent and overlap-

Ruhr-University of Bochum, Germany

D 2009 Massachusetts Institute of Technology

compared to upright human bodies. Human body shapes presented without heads yielded the opposite result pattern. The data for animal bodies did not yield consistent effects. Taken together, the present findings suggest that human bodies might be processed by specialized cortical mechanisms which are at least partly dissociable from mechanisms involved in object or face processing. &

ping but distinct networks within the fusiform gyrus (Peelen & Downing, 2005, 2007; Schwarzlose et al., 2005). This assumption is further supported by behavioral studies indicating that recognition of faces is more affected by stimulus inversion than recognition of nonface stimuli (inverted face effect). There is evidence of holistic (Farah, Tanaka, & Drain, 1995; Tanaka & Farah, 1993) as well as configural processing of human faces, and both are disrupted by inversion (Maurer et al., 2007; Leder, Candrian, Huber, & Bruce, 2001; Freire, Lee, & Symons, 2000). Reed, Stone, Grubb, and McGoldrick (2006) and Reed, Stone, Bozova, and Tanaka (2003) reported slower RTs and higher error rates for decisions about inverted compared to upright human body positions. Inversion effects did not emerge for isolated body parts or scrambled bodies (Reed et al., 2006). These data offer some support to the idea that human bodies—like faces—are processed configurally. Human faces elicit a negative event-related potential (ERP) peaking about 170 msec (termed N170) after stimulus onset, with maximum amplitudes in occipitotemporal areas (Minnebusch, Suchan, Ramon, & Daum, 2007; Eimer, 2000a, 2000b; Bentin, Deouell, & Soroker, 1999). N170 latencies are longer for inverted relative to upright faces; this effect is not observed for objects (Itier & Taylor, 2004a, 2004b; Eimer, 2000a; Rossion et al., 2000; Bentin, Allison, Puce, Perez, & McCarthy, 1996). N170 amplitudes are larger for inverted relative to upright faces (Marzi & Viggiano, in press; Itier & Taylor, 2004a, 2004b; Rossion et al., 1999, 2000). Thierry et al. (2006) described an ERP component evoked by stimuli representing the form of the human body (N190), which

Journal of Cognitive Neuroscience X:Y, pp. 1–10

STUDIEN

UND


differed significantly from the face-specific N170, with source localization identifying distinct sources for both components. Configural processing of human body shapes is supported by larger N190 amplitudes and longer latencies for inverted compared to upright human bodies (Stekelenburg & de Gelder, 2004). The latter study used photographs representing the whole body and masked the faces to minimize face processing. However, contextual cues (such as masked faces) can elicit object-specific neuronal responses (Cox, Meyers, & Sinha, 2004), and occipito-temporal activation evoked by body shapes is modulated by the presence or absence of the face (Morris, Pelphrey, & McCarthy, 2006). It is therefore possible that even masked faces might activate face-specific processing mechanisms and it is unclear whether bodies, faces, or both are critical for the reported effects. Taken together, the available evidence suggests that human bodies, like faces, are processed in specialized cortical areas. There is, however, as yet considerable uncertainty as to whether faces and bodies are processed by the same neuronal mechanisms (domain general hypothesis) or by dissociable mechanisms (face specificity hypothesis). The domain-general hypothesis states that the mechanisms engaged by faces are not specific for a particular stimulus category, but for a particular process (configural encoding), which is more important for recognizing faces than other stimulus classes (Tarr & Cheng, 2003). Processing of faces and bodies tends to be compared to processing of inanimate objects (Reed et al., 2003, 2006; Thierry et al., 2006; Peelen & Downing, 2005; Schwarzlose et al., 2005; Stekelenburg & de Gelder, 2004; Hole, George, & Dunsmore, 1999). Interestingly, the FFA/FBA was found to be strongly activated for animal bodies with heads (Downing et al., 2006; Chao, Martin, & Haxby, 1999; Kanwisher, Stanley, & Harris, 1999), but not for animals without heads (Kanwisher et al., 1999). The EBA yielded higher activations for pictures of mammals compared to birds, fishes, or objects (Downing et al., 2006), indicating that the EBA activations by animals with a body form similar to that of humans. It is as yet unknown whether inversion effects are specific for human bodies or whether they are also found for animal bodies. The aim of the present study was to further investigate whether inversion effects would support the hypothesis of configural processing of human bodies similar to the known effects for faces and whether processing of human bodies would differ from processing of animal bodies. It has previous been shown that EBA activity is modulated by the presence of a face (Morris et al., 2006). It is as yet unknown, however, whether the presence or absence of a face also modulates ERP amplitudes evoked by human body stimuli. To investigate this issue, we used human bodies with masked faces and headless human bodies as stimulus material.

2

Journal of Cognitive Neuroscience

METHODS Subjects Seventeen right-handed subjects (9 women, mean age = 23.8 years, SD = 3.9) participated in this study. All subjects had normal or corrected-to-normal vision. There was no history of neurological or psychiatric illness in any of the subjects. The study was performed in accordance with ethical standards laid down in the Declaration of Helsinki (Varga, 1975) and approved by the Ethics Committee of the Medical Faculty of the RuhrUniversity, Bochum, Germany. Written informed consent was obtained from all subjects.

Stimuli Gray -scaled pictures representing six different categories (human bodies, human bodies without head, birds, cats, dogs, and houses) were displayed on a white background. Pictures of houses were included to determine whether bodies generally evoked a larger early negative component than houses, similar to the facespecific N170. Each stimulus category entailed 60 items, at a size of approximately 38 � 38 of visual angle per item. Photographs for the human body categories were taken from a set of photographs of whole bodies of 30 men and 30 women in natural postures, which was developed in our department. A set of photographs of animals was selected from the Internet. Stimuli were matched with respect to body postures and orientation to control for within-category similarity, and the final stimulus selection was based on extensive piloting in healthy subjects. To minimize face processing during presentation of bodies, the faces on the pictures were masked (human bodies, birds, cats, and dogs). Consistent with the procedure described by previous investigations (Taylor et al., 2007; Downing et al., 2006; Peelen & Downing, 2005), a separate category of human bodies with heads removed was added to the stimulus material. The body stimuli (human bodies, human bodies without head, birds, cats, and dogs) were presented both in upright and inverted positions.

Procedure Subjects were seated in a sound-attenuating and electrically shielded room, facing a computer monitor at a distance of 80 cm. They were instructed to fixate the center of the screen and to avoid blinking or eye and body movements. The task was developed in analogy to the procedure described by Yovel and Kanwisher (2005). Pairs of bodies or houses were presented sequentially either in upright (all categories) or inverted positions (all categories except houses) in random order. In each trial, the first and second pictures—belonging to the same

Volume X, Number Y

STUDIEN

UND


stimulus category—were presented in the center of the monitor for 250 msec respectively, with an ISI of 500 msec and a mean ITI of 1000 msec (see Figure 1). Each trial lasted 2000 msec. During the ITI, a black exclamation mark appeared in the center of the monitor. The subjects’ task was to decide as quickly and accurately as possible whether the first and second stimuli were the same (same condition) or different (different condition). Same/different judgments were made via keypresses. Each stimulus was used once in the same and once in the different condition. Each of the 11 conditions (six categories; all categories except houses are presented upright and inverted) involved 60 trials in the ‘‘same condition’’ and 60 trials in the ‘‘different condition.’’ In total, 1320 stimuli were presented in three blocks of 440 trials each.

Data Analyses Behavioral Data Correct responses and reaction times (RTs) were assessed for each subject and each condition. Efficiency scores (see Jacques & Rossion, 2007) were calculated for each subject and each condition by dividing the mean RTs by the proportion of correct responses. This procedure was adopted to obtain a single parameter for both responses. Efficiency scores were also used in order to controls for possible speed–accuracy tradeoffs. A low score reflects good performance. For efficiency scores, a 5 � 2 � 2, repeated measures ANOVA with Greenhouse–Geisser corrections were performed, with factors stimulus category (human bodies, human bodies without head, birds, cats, and dogs), orientation (upright vs. inverted), and condition (same vs. different).

EEG Recordings Scalp recordings were obtained from 30 tin electrodes (10–20 standard set-up) mounted in an elastic cap. Four additional electrodes were placed above and below the left eye and on the outer canthus of each eye to record vertical and horizontal eye movements. The POz electrode served as reference. Electrode impedance was kept below 5 k� and digitized at a sampling rate of 250 Hz. EEG signals were filtered with a band-pass filter of 0.5–35 Hz. Trials with EOG or EEG artifacts exceeding 50 AV were omitted from further analyses.

EEG Data Analysis focused on the ERP response to the first picture of each pair (see Jacques & Rossion, 2007), for which both stimulus category and orientation were unpredictable, which was not the case for the second picture. All trials (correct and error trials) are included in the analysis. The raw data were segmented off-line in epochs of 550 msec, starting 200 msec prior to stimulus onset; activity 200 msec before stimulus onset served as baseline.

Figure 1. Experimental design.

Minnebusch, Suchan, and Daum

3

STUDIEN

UND


Table 1. Mean Percentage of Correct Responses and RTs (in milliseconds and SD in Brackets) for Each Stimulus Category and Orientation % Correct Responses Upright

Inverted

Median RT (msec) Upright

Inverted

Human body

92.4 (2.9)

90.3 (4.1)

511 (65)

519 (73)

Human body without head

92.7 (3.9)

92.4 (3.0)

519 (73)

518 (71)

Cat

93.2 (2.9)

92.9 (3.4)

510 (73)

508 (78)

Dog

93.3 (2.3)

93.2 (3.3)

504 (74)

517 (72)

Bird

92.5 (3.9)

90.3 (3.1)

509 (65)

506 (76)

ERPs were averaged separately for stimulus categories (human bodies, human bodies without head, birds, cats, dogs, and houses) and orientations (upright and inverted). The N170 peak amplitude was determined as the peak amplitude within the 140–200 msec poststimulus latency window relative to baseline at electrode positions P7/P8. Amplitude maxima were taken to calculate N170 latencies. Visual inspection suggested that the maximal amplitudes were seen at these positions in all conditions (see Figure 3). In the first analysis, N170 amplitudes for upright conditions were submitted to a 6 � 2 repeated measures ANOVA with factors stimulus category (human bodies, human bodies without head, birds, cats, dogs, and houses) and hemisphere (left vs. right) to investigate whether body categories evoked a significantly higher N170 than houses. N170 amplitudes and latencies were then submitted to repeated measures ANOVA with factors category (human bodies, human bodies without head, birds, cats, and dogs), orientation (upright vs. inverted), and hemisphere (left vs. right). Greenhouse–Geisser corrections were used where appropriate.

RESULTS Behavioral Results Correct responses and RTs are presented in Table 1. Efficiency scores for each condition are presented in Table 2. ANOVA yielded main effects of stimulus category [F(1, 17) = 4.4, p = .01], orientation [F(1, 17) = 6.9, p = .02], and condition [F(1, 17) = 16.2, p = .001], with better performance in the same compared to the different condition as well as an interaction between category and orientation [F(1, 17) = 4.4, p = .009; Figure 2]. Subsequent paired comparisons revealed a better performance for upright compared to inverted human bodies [F(1, 17) = 8.6, p = .01] and upright compared to inverted dogs [F(1, 17) = 6.3, p = .02; Figure 2]. Human bodies without heads showed the opposite pattern with better performance for inverted compared to upright bodies [F(1, 17) = 5.5, p = .03; Figure 2]. None of the other effects reached significance.

4


Electrophysiological Results Amplitude Grand averages elicited by upright human bodies, upright human bodies without head, upright birds, upright cats, upright dogs, and upright houses at left and right parietal electrodes (P7/P8) are presented in Figure 3. ANOVA yielded a main effect of category [F(1, 17) = 18.1, p < .001], with significantly smaller N170 amplitudes for houses compared to each of the five body categories (see Figure 3). A further ANOVA based on the body categories only (human bodies, human bodies without head, birds, cats, and dogs), and including the upright/inverted condition, yielded a main effect of category [F(1, 17) = 6.6, p = .001] and an interaction between category and orientation [F(1, 17) = 5.5, p = .006], as well as an Orientation � Hemisphere interaction [F(1, 17) = 11.9, p = .003], the latter being due to higher amplitudes for upright bodies compared to inverted bodies [F(1, 17) = 7.6, p = .02] in the left hemisphere. However, inverted human bodies evoked a larger N170 compared to upright human bodies [F(1, 17) = 6.3, p = .02]. Human bodies without heads [F(1, 17) = 5.3, p = .04] and cats [F(1, 17) = 5.8, p = .03] showed the Table 2. Efficiency Scores (SD in brackets) for Each Stimulus Category, Orientation, and Condition Category Human body

Human body without head Bird

Dog

Cat

Orientation

Same

Different

upright

531 (80.6)

577 (79.1)

inverted

540 (94.7)

614 (72.2)

upright

554 (104.6)

602 (76.0)

inverted

546 (91.4)

576 (70.0)

upright

527 (88.6)

574 (65.5)

inverted

532 (82.7)

593 (80.9)

upright

518 (83.8)

564 (81.6)

inverted

541 (95.6)

571 (71.8)

upright

520 (92.1)

573 (88.8)

inverted

529 (95.8)

574 (88.7)

Volume X, Number Y

STUDIEN

UND


Figure 2. (A) Efficiency scores and standard errors for upright and inverted human bodies and human bodies without heads. (B) Efficiency scores and standard errors for upright and inverted cats, dogs, and birds.

opposite pattern, with larger N170 for upright compared to inverted bodies (Figures 4 and 5). Latency A 5 � 2 � 2 ANOVA revealed a trend toward a significant interaction between category, orientation, and hemisphere [F(4, 17) = 2.6, p = .06]. Subsequent exploratory pairwise comparisons yielded longer N170 latencies for inverted (M = 174 msec, SD = 20) compared to upright human bodies (M = 168 msec, SD = 21) in the left hemisphere. None of the other effects reached significance.

DISCUSSION The aim of the present study was to further investigate the neurocognitive mechanisms underlying human

body perception by assessing the behavioral and electrophysiological effects of inversion of body stimuli. By including body forms without heads as a stimulus category, the findings should help to clarify whether human body forms, faces, or both are responsible for previously reported stimulus inversion effects on behavioral information processing and an early ERP component (N170). A further issue of interest was the specificity of these effects for human bodies relative to animal body forms. Consistent with previous findings (Reed et al., 2003, 2006; Stekelenburg & de Gelder, 2004), recognition of human bodies (whole shapes with heads) was significantly affected by inversion. The N170 amplitudes were higher and latencies were prolonged for inverted compared to upright human bodies. These results clearly suggest similarities to the well-known effects of inversion on processing of human aces, which has been linked to the perceptual encoding stage ( Jacques &


5

STUDIEN

UND


6


Volume X, Number Y

STUDIEN

UND


Figure 4. Grand mean maps for upright body categories (top). Grand-average ERPs elicited by upright and inverted body categories recorded at P7 (left side) and P8 (right side, bottom).

Rossion, 2007). As discussed by Rossion et al. (2000), a possible explanation for the enhanced and delayed N170 for inverted compared to upright faces implies that the effect reflects a mechanism specific for face perception. Processing of inverted faces recruits both face- and object-related regions in the brain. In addition, inverted faces are more difficult to process compared to upright faces and, therefore, recruit more processing resources. Previous reports of behavioral and electrophysiological body form inversion effects, using bodies with heads as stimuli (Reed et al., 2003, 2006; Stekelenburg & de Gelder, 2004), concluded that both configural and holistic processing might be important for the recognition of both human faces and bodies. However, in the present study, human bodies shown without heads yielded the opposite pattern to the face inversion effects, with better performance and a reduced N170 amplitude for inverted compared to upright human bodies. Comparable ERP inversion effects have also been reported for point-light walker stimuli, which provide information about human body shape but not the face ( Jokisch, Daum, Suchan, & Troje, 2005). There are two possible explanations for the observed effect. The results might

suggest that configural processing may be important for the recognition of human bodies with heads, representing a realistic silhouette, but not for bodies without heads. The latter stimulus class elicited longer N170 latencies compared to human bodies with heads. This implies that even masked faces (used as stimuli by Stekelenburg & de Gelder, 2004) might activate facespecific processing mechanisms to some degree, which in turn might be responsible for the reported behavioral and electrophysiological inversion effects in this study. Alternatively, if the N170 is a marker for configural processing that is elicited in response to both upright faces and upright human bodies with faces, the reversed inversion effect for human bodies without heads might be related to a lack of configural processing of these stimuli (comparable to inverted faces). Upright human bodies without heads may partly, but not closely, match the generalized template for human body shape and therefore provide a confusing stimulus input. Inverted human body shapes without heads, on the other hand, clearly do not match the typical representation. The present findings need to be interpreted with caution, and further investigations of face

Figure 3. Grand-average ERPs elicited by upright body categories (black lines) and upright houses recorded at P7 (left side) and P8 (right side).


7

STUDIEN

UND


Figure 5. (A) Grand-average amplitudes and standard errors at electrode positions P7 and P8 for upright and inverted human bodies and human bodies without heads. (B) Grand-average amplitudes and standard errors at electrode positions P7 and P8 for upright and inverted cats, dogs, and birds.

and body inversion effects and their implications for configural processing are needed. Because inversion disrupts configural processing, the reported inversion effect for human bodies with heads might suggest disruption of configural processing. However, it is not known which other processes are also disrupted by inversion and alternative explanations must remain open. A further explanation of the reversed effect for human bodies without heads would imply that the most distinctive feature of human bodies without heads is the position or shape of their feet, which are probably more prominent for inverted bodies without head. A previous study (Thierry et al., 2006) reported a body-sensitive ERP component peaking about 190 msec (N190) after stimulus onset. This observation was not replicated in the present study. Both human bodies with masked faces and bodies presented without heads elicited ERP components peaking about 170 msec after

8


stimulus onset, which are comparable to the well-known face-specific N170. Despite the fact that human bodies without heads yielded inversion effects (better recognition and reduced amplitude for inverted compared to upright human bodies), which are opposite to previously reported effects, processing of human bodies appears to be clearly dissociable from object processing. Recognition performance tends to be similar for upright and inverted objects (Reed et al., 2003, 2006) as are N170 amplitudes and latencies (Eimer, 2000a; Bentin et al., 1999; Tanaka & Farah, 1993). The present results are inconsistent with previously reported face and object perception data (Hole et al., 1999; Lewis & Johnston, 1997; Johnston, Hill, & Carman, 1992), which indicated that human bodies might not be processed configurally like faces or analytically like objects. There is evidence for clear differences between

Volume X, Number Y

STUDIEN

UND


the neuronal and functional systems involved in body and face processing (Peelen & Downing, 2005, 2007; Downing et al., 2001, 2006; Morris et al., 2006; Schwarzlose et al., 2005). Similarities in processing mechanisms of faces and human bodies have, however, also been reported (Reed et al., 2003, 2006; Stekelenburg & de Gelder, 2004). The present data indicate that human bodies, like faces, may represent a unique stimulus class with specialized processing mechanisms, which differ from face and object processing. All human and animal body categories elicited an ERP component peaking 170 msec after stimulus onset, which was larger for bodies than for houses. This component did not differ between human and animal bodies, supporting the assumption that the N170 is evoked by body forms in general. It should be noted that animal bodies were presented with heads in the present study. Recent imaging studies reported body-sensitive FBA activation for animal bodies with but not without heads (Downing et al., 2006; Chao et al., 1999; Kanwisher et al., 1999). It remains to be determined whether animal bodies without heads would activate an ERP component comparable to the N170 evoked by human bodies without heads. The N170 elicited by bodies is similar to the face-specific N170, indicating that faces and bodies might be processed by distinct but adjacent neuronal networks. For the animal categories, the behavioral and electrophysiological data show a heterogeneous pattern. The behavioral data indicate an inversion effect for dogs but not for cats and birds. Thus, inversion seems to influence the recognition of dogs more than the recognition of birds and cats, although there is as yet no convincing explanation for this finding. The electrophysiological data showed an enhanced N170 for upright cats compared to inverted cats but not for dogs and birds. These results are similar to those of human bodies without heads, tentatively suggesting that cats and human bodies without heads might share some processing mechanisms. There is as yet no explanation for this puzzling and unexpected pattern of results. The underlying mechanisms remain to be explored in further studies, which should include human faces as well as animal bodies with and without heads to compare directly the well-known face inversion effect with inversion effects for other stimuli. In summary, there is no clear evidence for an exclusively configural or analytical processing mechanism involved in human body form perception. A surprising finding is the better recognition of inverted compared to upright body shapes without heads, which might be related to a lack of configural processing of these stimuli. In everyday life, human body shapes are always perceived with heads. Pictures of human bodies without heads might be disturbing stimuli with negative connotations, which lead to higher error rates and slower processing. In summary, human bodies seem to be a special

stimulus class, which are processed by specialized neuronal areas which are at least partly dissociable from areas responsible for face or object processing.

UNCITED REFERENCES Gauthier, Behrmann, & Tarr, 1999 Gauthier, Tarr, Anderson, Skudlarski, & Gore, 1999 Tarr & Gauthier, 2000 Acknowledgments This work was supported by the Ruhr-University Research School founded by the DFG in the framework of the Excellence Initiative. We thank Petra Jordan for her help during the data collection. Reprint requests should be sent to Denise A. Minnebusch, Institute of Cognitive Neuroscience, Department of Neuropsychology, Ruhr-University of Bochum, Universita¨tsstraße 150, D-44780 Bochum, Germany, or via e-mail: Denise.Minnebusch@ rub.de.

REFERENCES Bentin, S., Allison, T., Puce, A., Perez, E., & McCarthy, G. (1996). Electrophysiological studies of face perception in humans. Journal of Cognitive Neuroscience, 8, 551–565. Bentin, S., Deouell, L. Y., & Soroker, N. (1999). Selective visual streaming in face recognition: Evidence from developmental prosopagnosia. NeuroReport, 10, 823–827. Chao, L. L., Martin, A., & Haxby, J. V. (1999). Are faceresponsive regions selective only for faces? NeuroReport, 10, 2945–2950. Cox, D., Meyers, E., & Sinha, P. (2004). Contextually evoked object-specific responses in human visual cortex. Science, 304, 115–117. Downing, P. E., Chan, A. W., Peelen, M. V., Dodds, C. M., & Kanwisher, N. (2006). Domain specificity in visual cortex. Cerebral Cortex, 16, 1453–1461. Downing, P. E., Jiang, Y., Shuman, M., & Kanwisher, N. (2001). A cortical area selective for visual processing of the human body. Science, 293, 2470–2473. Eimer, M. (2000a). Effects of face inversion on the structural encoding and recognition of faces. Evidence from event-related brain potentials. Brain Research, Cognitive Brain Research, 10, 145–158. Eimer, M. (2000b). Event-related brain potentials distinguish processing stages involved in face perception and recognition. Clinical Neurophysiology, 111, 694–705. Farah, M. J., Tanaka, J. W., & Drain, H. M. (1995). What causes the face inversion effect? Journal of Experimental Psychology: Human Perception and Performance, 21, 628–634. Freire, A., Lee, K., & Symons, L. A. (2000). The face-inversion effect as a deficit in the encoding of configural information: Direct evidence. Perception, 29, 159–170. Gauthier, I., Behrmann, M., & Tarr, M. J. (1999). Can face recognition really be dissociated from object recognition? Journal of Cognitive Neuroscience, 11, 349–370. Gauthier, I., Tarr, M. J., Anderson, A. W., Skudlarski, P., & Gore, J. C. (1999). Activation of the middle fusiform ‘‘face area’’ increases with expertise in recognizing novel objects. Nature Neuroscience, 2, 568–573.


9

STUDIEN

UND


Gauthier, I., Tarr, M. J., Moylan, J., Skudlarski, P., Gore, J. C., & Anderson, A. W. (2000). The fusiform ‘‘face area’’ is part of a network that processes faces at the individual level. Journal of Cognitive Neuroscience, 12, 495–504. Haxby, J. V., Hoffman, E. A., & Gobbini, M. I. (2000). The distributed human neural system for face perception. Trends in Cognitive Sciences, 4, 223–233. Haxby, J. V., Ungerleider, L. G., Clark, V. P., Schouten, J. L., Hoffman, E. A., & Martin, A. (1999). The effect of face inversion on activity in human neural systems for face and object perception. Neuron, 22, 189–199. Hole, G. J., George, P. A., & Dunsmore, V. (1999). Evidence for holistic processing of faces viewed as photographic negatives. Perception, 28, 341–359. Itier, R. J., & Taylor, M. J. (2004a). Effects of repetition learning on upright, inverted and contrast-reversed face processing using ERPs. Neuroimage, 21, 1518–1532. Itier, R. J., & Taylor, M. J. (2004b). Face recognition memory and configural processing: A developmental ERP study using upright, inverted, and contrast-reversed faces. Journal of Cognitive Neuroscience, 16, 487–502. Jacques, C., & Rossion, B. (2007). Early electrophysiological responses to multiple face orientations correlate with individual discrimination performance in humans. Neuroimage, 36, 863–876. Johnston, A., Hill, H., & Carman, N. (1992). Recognising faces: Effects of lighting direction, inversion, and brightness reversal. Perception, 21, 365–375. Jokisch, D., Daum, I., Suchan, B., & Troje, N. F. (2005). Structural encoding and recognition of biological motion: Evidence from event-related potentials and source analysis. Behavioural Brain Research, 157, 195–204. Kanwisher, N., McDermott, J., & Chun, M. M. (1997). The fusiform face area: A module in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17, 4302–4311. Kanwisher, N., Stanley, D., & Harris, A. (1999). The fusiform face area is selective for faces not animals. NeuroReport, 10, 183–187. Kanwisher, N., & Yovel, G. (2006). The fusiform face area: A cortical region specialized for the perception of faces. Philosophical Transactions of the Royal Society of London, Series B, Biological Sciences, 361, 2109–2128. Leder, H., Candrian, G., Huber, O., & Bruce, V. (2001). Configural features in the context of upright and inverted faces. Perception, 30, 73–83. Lewis, M. B., & Johnston, R. A. (1997). The Thatcher illusion as a test of configural disruption. Perception, 26, 225–227. Marzi, T., & Viggiano, M. P. (in press). Interplay between familiarity and orientation in face processing: An ERP study. International Journal of Psychophysiology. Maurer, D., O’Craven, K. M., Le Grand, R., Mondloch, C. J., Springer, M. V., Lewis, T. L., et al. (2007). Neural correlates of processing facial identity based on features versus their spacing. Neuropsychologia, 45, 1438–1451. Minnebusch, D. A., Suchan, B., Ramon, M., & Daum, I. (2007). Event-related potentials reflect heterogeneity of developmental prosopagnosia. European Journal of Neuroscience, 25, 2234–2247. Morris, J. P., Pelphrey, K. A., & McCarthy, G. (2006). Occipitotemporal activation evoked by the perception of

10


human bodies is modulated by the presence or absence of the face. Neuropsychologia, 44, 1919–1927. Peelen, M. V., & Downing, P. E. (2005). Selectivity for the human body in the fusiform gyrus. Journal of Neurophysiology, 93, 603–608. Peelen, M. V., & Downing, P. E. (2007). The neural basis of visual body perception. Nature Reviews Neuroscience, 8, 636–648. Peelen, M. V., Wiggett, A. J., & Downing, P. E. (2006). Patterns of fMRI activity dissociate overlapping functional brain areas that respond to biological motion. Neuron, 49, 815–822. Reed, C. L., Stone, V. E., Bozova, S., & Tanaka, J. (2003). The body-inversion effect. Psychological Science, 14, 302–308. Reed, C. L., Stone, V. E., Grubb, J. D., & McGoldrick, J. E. (2006). Turning configural processing upside down: Part and whole body postures. Journal of Experimental Psychology: Human Perception and Performance, 32, 73–87. Rossion, B., Delvenne, J. F., Debatisse, D., Goffaux, V., Bruyer, R., Crommelinck, M., et al. (1999). Spatio-temporal localization of the face inversion effect: An event-related potentials study. Biological Psychology, 50, 173–189. Rossion, B., Gauthier, I., Tarr, M. J., Despland, P., Bruyer, R., Linotte, S., et al. (2000). The N170 occipito-temporal component is delayed and enhanced to inverted faces but not to inverted objects: An electrophysiological account of face-specific processes in the human brain. NeuroReport, 11, 69–74. Schwarzlose, R. F., Baker, C. I., & Kanwisher, N. (2005). Separate face and body selectivity on the fusiform gyrus. Journal of Neuroscience, 25, 11055–11059. Stekelenburg, J. J., & de Gelder, B. (2004). The neural correlates of perceiving human bodies: An ERP study on the body-inversion effect. NeuroReport, 15, 777–780. Tanaka, J. W., & Farah, M. J. (1993). Parts and wholes in face recognition. Quarterly Journal of Experimental Psychology, Series A, 46, 225–245. Tarr, M. J., & Cheng, Y. D. (2003). Learning to see faces and objects. Trends in Cognitive Sciences, 7, 23–30. Tarr, M. J., & Gauthier, I. (2000). FFA: A flexible fusiform area for subordinate-level visual processing automatized by expertise. Nature Neuroscience, 3, 764–769. Taylor, J. C., Wiggett, A. J., & Downing, P. E. (2007). fMRI analysis of body and body part representations in the extrastriate and fusiform body areas. Journal of Neurophysiology, 98, 1626–1633. Thierry, G., Pegna, A. J., Dodds, C., Roberts, M., Basan, S., & Downing, P. E. (2006). An event-related potential component sensitive to images of the human body. Neuroimage, 32, 871–879. Urgesi, C., Berlucchi, G., & Aglioti, S. M. (2004). Magnetic stimulation of extrastriate body area impairs visual processing of nonfacial body parts. Current Biology, 14, 2130–2134. Varga, A. C. (1975). Declaration of Helsinki (Adopted by the 18th World Medical Assembly in Helsinki, Finland, and revised by the 29th World Medical Assembly in Tokyo, 1975). In The main issue in bioethics (revised ed.). New York: Paulist Press. Yovel, G., & Kanwisher, N. (2005). The neural basis of the behavioral face-inversion effect. Current Biology, 15, 2256–2262.

Volume X, Number Y

STUDIEN

UND


Studie 3

Gradual inversion effects indicate configural processing of human body shapes Denise A. Minnebusch, Philipp M. Keune, Boris Suchan und Irene Daum, in Vorbereitung

STUDIEN

UND


Gradual inversion effects indicate configural processing of human body shapes Denise A. Minnebusch, Philipp M. Keune, Boris Suchan & Irene Daum, in Vorbereitung

Institute of Cognitive Neuroscience, Dept. of Neuropsychology, Ruhr-University of Bochum, Germany Corresponding author: Denise A. Minnebusch

STUDIEN

UND


Abstract It is as yet unclear whether the perception of human body forms, like the perception of human faces, is based on configural processing mechanisms. Configural face processing is reflected in the face inversion effect; inverted faces are disproportionately more difficult to recognize than inverted objects and this effect correlates with an electrophysiological inversion effect. The present study aimed to investigate the course of behavioral and electrophysiological body inversion effects depending upon the degree of deviation from the upright position (0°, 360°) to the inverted position (180˚). Body stimuli were presented either with heads (masked face) or without heads in a delayed matching task. For human bodies presented with and without heads, there was a quadratic relationship between the angle of rotation and the behavioral performance as well as the N170 amplitude, with maximum performance disruption at 180° deviation from the upright stimulus. The data indicate that configural body processing occurs during the structural encoding of body stimuli. In addition, the results support the view of parallel processing mechanisms for faces and bodies.

STUDIEN

UND


Introduction Previous research on person perception has focussed predominately on the mechanisms and neuronal correlates of face perception (Kanwisher & Yovel, 2006; Tsao & Livingstone, 2008). However, in recent years research interest into how body shapes are processed by the human brain has started to emerge (for review, see Peelen & Downing, 2007), and both similarities and differences between face and body perception have been described. It is well known that face perception activates specialized mechanisms, which analyze the configuration of the face rather than single face features (Leder & Bruce, 2000; Maurer et al., 2002). Configural processing is defined as perception based on relations among the features of a stimulus, and face perception involves configural as well as feature-based processing mechanisms. Other stimulus classes (e.g., houses, cars) appear to be analyzed on the basis of single features (Collishaw & Hole, 2000; Leder & Bruce, 2000; Maurer et al.,2002; Tanaka & Farah, 1993). Critical evidence for configural face processing comes from the face inversion effect, first described by Yin (1969; 1970): Faces presented upside down are more difficult to recognize than uprightly presented faces; this effect is more pronounced for faces compared to other stimuli. The face inversion effect is a robust phenomenon and has been replicated in a range of studies (for review, see Rossion & Gauthier 2002; Taylor et al., 2004). By studying the inversion effect, Reed et al. (2003; 2006) provided evidence that configural processing might be also important for human body shape processing, given faster reaction times (RTs) and better behavioral performances for uprightly presented compared to inverted bodies. Evidence supporting the assumption that both faces and bodies are processed differently than other stimulus classes stems from functional neuroimaging studies. Regions in the fusiform gyrus (termed fusiform face area, FFA) and the occipital lobe (termed occipital face area, OFA) showed disproportionate activations associated with faces compared to other stimuli (Gauthier et al., 2000; Haxby et al., 2000; Kanwisher et al., 1997). Areas which appear to be specific for body processing are the extrastriate body area (EBA) and the fusiform body area (FBA; Downing et al., 2001; 2007; Peelen & Downing, 2005; Schwarzlose et al., 2005). The EBA responds selectively to static and dynamic human bodies and body parts (Peelen & Downing, 2007). The FBA shows more pronounced activation for whole bodies compared to body parts (Taylor et al., 2007). Interestingly, the FBA is adjacent to and partly overlaps with the FFA (Peelen & Downing 2005; Schwarzlose et al., 2005). Studies investigating the face inversion effect showed higher activations in the FFA for uprightly presented compared to inverted faces, and this effect correlated with the behavioral face inversion effect (Kanwisher et al., 1998; Yovel & Kanwisher, 2005). Comparable fMRI studies examining the inversion effect for human body forms are not available as yet.

STUDIEN

UND


Faces elicit a negative event-related potential (ERP) peaking about 170 ms (N170) after stimulus onset (Bentin et al., 1996; Eimer, 2000b; Minnebusch et al., 2007). Inverted faces elicited longer N170 latencies and higher amplitudes compared to upright faces (Itier & Taylor, 2004b; Rossion et al., 1999; 2000). The N170 is associated with brain processes involved in late stages of structural encoding, where representations of global face configurations are generated for face recognition (Eimer, 2000b; Jacques & Rossion, 2007). Perception of human faces also affects an early endogenous ERP component, around 100-120 ms after stimulus onset (termed P100). Faces elicited an enhanced P100 compared to other stimuli. In addition, the P100 is enhanced and delayed for inverted compared to upright faces (Itier & Taylor, 2004a; c; Rossion et al.,1999; 2000). The P100 seems to reflect the perception of a stimulus as a face (Herrmann et al., 2005; Itier & Taylor, 2004c). Jacques and Rossion (2007) investigated the time window in which processing differences between upright and inverted faces start to emerge, i.e., the time window for the onset of configural processing. They presented faces at different orientations (from 0° to 360° in steps of 30°) to study the effects on the P100 and the N170 as well as performance on a matching task as a function of the angle of face rotation. Jacques and Rossion (2007) found a quadratic relationship between the angles of rotation, behavioral performance and both the P100 and the N170 amplitude. Behavioral performance did not correlate with the ERP signal before the N170 window. As the N170 reflects mechanisms associated with the structural encoding of a stimulus, these data provide evidence for the assumption that the behavioral face inversion effect occurs during perceptual face encoding. Electrophysiological studies have identified several ERP components associated with body perception. Human bodies evoke a characteristic positive component approximately 100 ms after stimulus onset (P100; Thierry et al., 2006) and a negative component at about 190 ms (N190) after stimulus onset (Kovacs et al., 2006; Stekelenburg & de Gelder, 2004; Thierry et al., 2006), similar to the N170 obtained for faces. The N190 generalizes to some degree to schematic descriptions of the human body form, and the source of the N190 differed from that of the N170 in the right posterior extrastriate cortex (Thierry et al., 2006). The electrophysiological inversion effect for bodies was characterized by longer latencies and higher amplitudes of the N190 for inverted compared to upright stimuli (Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004). Body inversion did not have a significant effect on ERPs until about N190 after stimulus onset (Righart & de Gelder, 2007). Both studies administered body stimuli which included the head, even though the faces were masked to minimize the involvement of face processing. Occipitotemporal activation, however, has been shown to vary according to the presence or absence of a face (Morris et al., 2006) and even the shape of the head may serve as a contextual cue for the activation of specialized mechanisms which induce configural processing (Cox et al., 2004). For bodies without heads, the inversion effect itself may be reversed, with behavioral data indicating better performance on recognition tasks for inverted compared to upright

STUDIEN

UND


bodies and a similar inversion emerged for the N170 amplitude (Minnebusch et al., 2009). These findings cast doubt on a simple mapping of face inversion to body shape inversion effects. The fact that inverted headless bodies appear to be processed more efficiently than upright headless bodies does not lend itself to a straightforward interpretation. Contextual cues delivered by the shape of the head appear to play an important role in eliciting configural processing for body shapes. Headless bodies might not sufficiently match the general template of bodies to induce configural processing. An idea that complements this notion is that the most salient feature of headless bodies are the feet, which might be the major basis for decisions in recognition tasks (Minnebusch et al., 2009). Body parts do not produce inversion effects (Reed et al., 2006) and focusing on the feet of a body implies feature-based processing, which would be more prominent for inverted headless bodies. The present study aimed to investigate the course of behavioral and electrophysiological body inversion effects depending upon the degree of deviation from the 0˚ (360˚) toward the 180˚ position in steps of 45°. Body stimuli were presented either with heads (masked face) or without heads in order to determine the effects of the presence of a head at different rotation angles. Comparable to previously reported effects of gradual face rotation (Jacques & Rossion, 2007), we expected a quadratic relationship between the degree of rotation and the behavioral and the electrophysiological data.

Method Subjects Sixteen right-handed subjects (8 female, mean age 26.8 years, SD = 2.3) participated in the present study. Subjects had normal or corrected-to-normal vision and all participants reported normal neurological and psychiatric health. The study was performed in accordance with the ethic standards laid down in the declaration of Helsinki (Varga 1975) and was approved by the Ethic Committee of the of Medical Faculty of the Ruhr-University, Bochum, Germany. Written informed consent was obtained from all subjects.

Stimuli The stimuli were gray-scaled front view photographs of 60 people (30 female) wearing dark clothing in natural postures. Bodies extended over approximate 3˚ visual angle in height and were displayed on a

STUDIEN

UND


white background. Following the procedure described by Minnebusch et al. (2009), two categories of stimuli were included, bodies with heads (masked face) and bodies without heads, in order to keep face processing at a minimum. Each original photograph was edited to yield a version for each stimulus class. The body stimuli of each category were presented at eight different angles (0˚, 45˚, 90˚, 135˚, 180˚, 225˚, 270˚, and 315˚) with 60 pairs per angle for each stimulus category.

Procedure Participants were seated in a light- and sound-attenuated room at a distance of 80 cm from a computer screen. The matching task was based on the procedure described by Minnebusch et al. (2009). In each trial, two body pictures were presented sequentially, rotation angles were randomized across trials. A trial started with the presentation of the first body picture for 250 ms. After an inter stimulus interval of 500 ms, the second body picture was displayed for 250 ms followed by an exclamation mark for 1000 ms which was presented at the center of the screen, resulting in a total trial length of 2000 ms. Note that category and rotation angle of the two bodies within a pair were identical. An illustration of the design and stimulus orientations is provided in Figure 1. The participants’ task was to decide via a key press whether the first and the second stimuli were the same (same condition) or different (different condition). Instructions stressed both speed and accuracy. In sum, each subject completed 960 trials, separated into two blocks of approximately 15 minutes. For each angle, 120 trials were presented (60 per stimulus category, i.e., bodies with and without heads). For each stimulus category, 30 trials represented the same and 30 the different condition, respectively.

STUDIEN

UND


Figure 1: Example of the stimulus classes (a) and experimental design (b).

STUDIEN

UND


Electroencephalography recordings Electroencephalography (EEG) was recorded via 30 electrodes mounted in an electrode cap according to the 10-20 standard setup. Vertical and horizontal eye movements were recorded via four additional electrodes, with two electrodes placed above and below the right eye and two placed on the lateral canthi of the eyes. Electrode impedance was kept below 5 kΩ. All active electrodes were referenced to two mastoid electrodes. EEG was digitized at a sampling rate of 250 Hz and signals were filtered with a bandpass filter of 0.5-35 Hz. Trials with eye movements or EEG artifacts exceeding 50 µV were excluded.

Data analysis Behavioral data Percentage of correct responses and RTs were assessed for each subject and each condition. Correct responses and RTs were submitted to repeated measures ANOVA with Category (bodies with head vs. bodies without head) and Orientation (8 angles) as within-subject factors. Greenhouse-Geisser corrections were used when appropriate. Orientation effects were analyzed in terms of orthogonal polynominal contrasts.

EEG data As in previous studies (Jacques & Rossion, 2007; Minnebusch et al., 2009), analysis focused on the ERP response to the first picture of each pair for which both stimulus category and orientation were unpredictable. Correct and error trials are included in the analysis. The raw data were segmented offline in epochs of 550 ms, starting 200 ms prior to stimulus onset, activity 100 ms before stimulus onset served as baseline. ERPs were averaged separately for both stimulus categories (bodies with heads, bodies without heads) and all orientations (0˚, 45˚, 90˚, 135˚, 180˚, 225˚, 270˚, and 315˚). The relevant components were the P100 and the N170. P100 amplitude was determined as the peak amplitude between 80 and 140 ms after stimulus onset on electrode positions P7/P8 and PO7/PO8. The N170 peak amplitude was assessed as the peak amplitude within the 140-200 ms post-stimulus latency window relative to baseline at electrode positions P7/P8 and PO7/PO8. Amplitude maxima were used to determine P100 and N170 latencies.

STUDIEN

UND


Visual inspection suggested that the maximal amplitudes emerged at these electrode positions in both conditions. P100 and N170 amplitudes and latencies were submitted to repeated-measures ANOVA with Category (bodies with heads vs. bodies without heads), Orientation (0˚, 45˚, 90˚, 135˚, 180˚, 225˚, 270˚, and 315˚), Electrode position (P7/P8 vs. PO7/PO8) and Hemisphere (right vs. left) as within-subject factors. Orientation effects were analyzed based on polynominal contrasts.

Correlation analysis To further assess rotation effects, Pearson’s correlation coeffficients were determined between behavioral performance (percentage of correct responses and RTs) and the P100 and N170 amplitudes. Bonferroni correction for multiple comparisons was used. Separate correlations were analyzed for each category and each behavioral measure. To remove interindividual differences unrelated to the task, z-score transformations for the behavioral and electrophysiological data for each subject and orientation entered analysis.

Results Behavioral data Behavioral performance (percentage correct responses, RTs) for the different body orientations are illustrated in Figure 2. For correct responses, analysis revealed a main effect of Category (F1,15 = 24.1, p < 0.001), with better performance for bodies with heads compared to bodies without heads and a significant Orientation effect (quadratic trend: F1,15 = 11.3, p = 0.004; linear trend: F1,15 = 5.3, p = 0.036 ). Analysis of RTs yielded a significant effect of Orientation (quadratic trend: F1,15 = 9.9, p = 0.007). None of the other comparisons reached significance. In summary, the behavioral data indicate poorest performance around 180° rotation, with gradual improvement towards the 0° and 360° end points.

STUDIEN

UND


Figure 2: Percentage of correct responses and response times (with standard errors) as a function of body orientation. The 360° point is a duplicate of the 0° point.

STUDIEN

UND


Electrophysiological results P100 amplitude: Grand average amplitudes elicited by human bodies of different orientations are illustrated in Figure 3. ANOVA yielded significant Electrode position (F1,15 = 26.9, p < 0.001) and Hemisphere effects (F1,15 = 9.6, p = 0.008). P100 amplitude was higher at electrode positions P07/PO8 relative to P7/P8 and in the right hemisphere compared to the left hemisphere (see Figure 3). The Orientation effect was also significant, with both significant linear (F1,15 = 17.7, p = 0.001) and quadratic trends (F1,15 = 7.2, p = 0.018). None of the other comparisons reached significance.

Figure 3: P100 amplitude and standard errors at electrode position P7/P8 (a) and PO7/PO8 (b) as a function of body orientation. The 360° point is a duplicate of the 0° point.

STUDIEN

UND


P100 latency: For P100 latency, significant Orientation x Hemisphere (linear trend: F1,15 = 28.4, p < 0.001) and Category x Orientation x Hemisphere interactions emerged (linear trend: F1,15 = 28.4, p < 0.001). To resolve the three-way interaction, separate ANOVAs were performed for the two categories (human bodies with and without heads). For bodies with heads, none of the comparisons reached significance. For bodies without heads, we observed a significant Orientation x Hemisphere interaction (linear trend: F1,15 = 80.5, p < 0.001). The relationship between body orientation and P100 latency in both hemispheres was linear (left hemisphere: F1,15 = 34.0, p < 0.001; right hemisphere: F1,15 = 13.8, p = 0.003). N170 amplitude: N170 amplitudes elicited by bodies at different orientations at electrode positions P7/P8 and PO7/PO8 are presented in Figure 4. Analysis revealed a significant effect of orientation (linear trend: F1,15 = 5.1, p = 0.04; quadratic trend: F1,15 = 13.8, p = 0.002), an interaction between Orientation and Electrode position (F1,15 = 4.7, p = 0.003), an interaction between Orientation and Hemisphere (linear trend: F1,15 = 20.3, p < 0.001), and a three-way interaction between Orientation, Electrode position and Hemisphere (quadratic trend: F1,15 = 7.8, p = 0.014). To resolve the three-way interaction, separate ANOVAs were conducted for the Electrode positions (P7/P8 and PO7/PO8). For P7/P8, ANOVA yielded a significant Orientation effect (linear trend: F1,15 = 7.5, p = 0.016; quadratic trend: F1,15 = 13.8, p = 0.002), and an Orientation x Hemisphere interaction (linear trend: F1,15 = 19.5, p = 0.001). In both hemispheres, polynomial contrasts revealed that the relationship between N170 amplitude and body orientation was a combination of a linear (P7: F1,15 = 6.7, p = 0.02 , P8: F1,15 = 22.1, p < 0.001) and a quadratic component (P7: F1,15 = 10.1, p = 0.007, P8: F1,15 = 6.0, p = 0.03; see Figure 4). For PO7/PO8, a significant Orientation effect (quadratic trend: F1,15 = 12.1, p = 0.004) and a significant Orientation x Hemisphere interaction (linear trend: F1,15 = 15.8, p = 0.001) emerged. The orientation effect followed a quadratic trend in the left (F1,15 = 15.0, p = 0.002) and a linear trend in the right hemisphere (F7,15 = 18.2, p = 0.001; see Figure 4).

STUDIEN

UND


Figure 4: N170 amplitude and standard errors at electrode position P7/P8 (a) and PO7/PO8 (b) as a function of body orientation. The 360° point is a duplicate of the 0° point.

STUDIEN

UND


N170 latency: ANOVA revealed main effects of Category (F 1,15 = 4.6, p = 0.049) and Electrode position (F 7,15 = 4.2, p = 0.01) and a significant Orientation x Hemisphere interaction (linear trend: F 1,15 = 6.6, p = 0.022), as well as a significant Category x Orientation x Hemisphere interaction (linear trend: F1,15 = 12.4, p = 0.003). N170 latency was longer for bodies without heads compared to bodies with heads and longer at electrode position PO7/PO8 compared to electrode position P7/P8. To resolve the three-way interaction, separate ANOVAs were performed for the two categories (bodies with and without heads). For bodies with heads, none of the comparisons reached significance. For bodies without heads, ANOVA yielded a significant Orientation x Hemisphere interaction (linear trend: F1,15 = 16.5, p = 0.001). N170 latency followed a linear trend with body orientation in the left hemisphere (linear trend: F 1,15 = 9.7, p = 0.007).

Correlation analysis Significant correlations between behavioral performance and ERP components were mainly observed for N170 amplitudes. Correlations significant at the 0.006 level were considered. Correct responses: For bodies with heads, significant correlations emerged between percentage of correct responses and N170 amplitude at electrode positions P7 (r = 0.39, p < 0.001), P8 (r = 0.36, p < 0.001) and PO7 (r = 0.29, p = 0.001). Percentage of correct responses correlated negatively with P100 amplitude at electrode position P8 (r = -0.25, p = 0.005). For human bodies without heads, the correct response rates correlated significantly with N170 amplitude at electrode position P8 (r = 0.33, p < 0.001). Reaction times: For bodies without heads, RTs correlated negatively with N170 amplitude at electrode positions P7 (r = -0.22, p = 0.016) and P8 (r = -0.26, p = 0.004). In addition, significant correlations were observed for bodies with heads at PO8 (r = -0.29, p = 0.001). In the P100 time window, RTs for bodies with heads correlated positively with the PO8 amplitude (r = 0.26, p = 0.005).

Discussion The current study aimed to investigate the course of the body inversion effect with increasing deviation from 0° (360°) towards 180° rotation angle. Behavioral performance might be unaffected until a certain

STUDIEN

UND


degree of rotation has been reached and decline abruptly at a specific rotation angle. Alternatively, body rotation might gradually affect recognition performance. A further issue of interest was the relationship between early ERP components (P100 and N170) and the behavioral rotation effect. Finally, the study aimed at clarifying previously reported processing differences between bodies presented with and without heads (Minnebusch et al., 2009). Body rotation affected both behavioral and electrophysiological measures of body processing. The effects followed a quadratic trend, indicating a systematic relationship between the degree of deviation from the upright to the inverted representation (see Figures 2-4). In line with Jacques and Rossion (2007), who investigated the effect of face rotation, the later N170 and not the earlier P100 correlated with the behavioral rotation effect. This suggests a functional dissociation between the P100 and the N170, with the N170 being a component genuinely associated with face processing, while the P100 might be more sensitive to low-level differences such as the location of contrast within a given stimulus. This hypothesis is supported by the fact that the P100 is the earliest index of endogenous low-level feature processing of visual stimuli (Mangun, 1995). The P100 correlates with the degree of noise in an image (Tarkiainen et al., 2002), but not with the amount of face information (Jemel et al., 2003). The results of the present study indicate that the suggested dissociation of P100 and N170 might be generalized to the stimulus category of human bodies. For both the P100 and N170, the effect of body rotation followed both significant linear and quadratic trends, with evidence of a stronger linear trend for the P100 and a stronger quadratic trend for the N170. The current data also suggest that the behavioral body inversion effect is reflected by the N170 time window. As the N170 is thought to be generated by brain processes involved in the structural encoding of a stimulus (Eimer, 2000a; b), the body inversion effect appears to be associated with the structural encoding of a stimulus as a body. The present results are consistent with previously reported similarities between faces and bodies: Both categories activate comparable and partly overlapping areas in the fusiform gyrus (Schwarzlose et al., 2005), undergo configural processing (Maurer et al., 2002; Reed et al., 2003; 2006; Yin, 1969; 1970), and elicit similar brain potentials (Bentin et al., 1996; Stekelenburg & de Gelder, 2004). The present data also suggest that configural body processing, like configural face processing (Jacques & Rossion, 2007), starts to occur 170 ms after stimulus onset. The quadratic relationship between the degree of rotation and the behavioral performance was reported for human faces (Jacques & Rossion, 2007), and we observed a comparable pattern for bodies. Despite these similarities, there is also evidence for differences in the underlying mechanisms of face and body perception: First, the neuronal source of the face-sensitive N170 can be distinguished from the body-sensitive N170 (Peelen & Downing, 2005; Stekelenburg & de Gelder, 2004). Second, several

STUDIEN

UND


studies reported clear dissociations of processing areas for faces, bodies and objects in the extrastriate cortex (Downing et al., 2001; 2007; Gauthier et al., 2000; Peelen & Downing, 2007; Pitcher et al., 2009). Finally, the lack of a selective deficit in body processing comparable to a selective face processing deficit (termed prosopagnosia; Bodamer, 1947) supports the view that bodies and faces are processed by similar but dissociable cortical mechanisms. Moro et al. (2008) described a condition termed body form agnosia, which was associated with lesions which included damage to the EBA (Moro et al., 2008). However, body processing deficits were assessed using stimuli representing body parts but not with whole body forms (Moro et al., 2008). There are also no data on general object processing abilities provided for these subjects. Further studies are clearly needed to clarify whether focal lesions in the EBA or FBA cause a selective body processing impairment. In sum, there is evidence for both similarities and differences between face and body perception. The current study yielded a quadratic relationship between upright and inverted representations of body shapes and the N170 and behavioral performance on a matching task. The N170 was significantly related to behavioral performances. Comparable relationships were previously reported for faces (Jacques & Rossion, 2007), supporting the assumption that overlapping mechanisms are recruited for the processing of human faces and bodies. In contrast to a recent study which reported opposite inversion effects for bodies with and without heads (Minnebusch et al., 2009), rotation effects were similar for both types of stimuli. However, similar to our previous data (Minnebusch et al., 2009), performance was generally better for bodies with heads compared to bodies without heads, indicating that naturalistic body representations (bodies with heads) are easier to recognize than unnatural representations (bodies without heads). The ERP latency difference might be attributed to the unusual representation of bodies without heads. Despite these differences, body rotation clearly had similar effects on both stimulus categories, supporting the assumption that the effects are body specific and not simply caused by the presence or absence of the head. In summary, similar to configural face processing, configural body processing was linked to the N170 time window. The relationship between body rotation and behavioral responses followed a quadratic trend, as did the relationship between body orientation and the N170. Taken together, the general pattern of findings indicates that configural body processing occurs during structural encoding of body shapes (see Figure 4). The relationship between body rotation and the P100, on the other hand, followed a different trend (see Figure 3).

STUDIEN

UND


ReferenceS Bentin S, Allison T, Puce A, Perez E, McCarthy G. Electrophysiological studies of face perception in humans. J.Cogn.Neurosci. 1996; 8: 551-565. Bodamer J. Die Prosopagnosie. Archiv für Psychiatrie und Nervenkrankheiten 1947; 179: 6-53. Collishaw SM, Hole GJ. Featural and configurational processes in the recognition of faces of different familiarity. Perception 2000; 29: 893-909. Cox D, Meyers E, Sinha P. Contextually evoked object-specific responses in human visual cortex. Science 2004; 304: 115-117. Downing PE, Jiang Y, Shuman M, Kanwisher N. A cortical area selective for visual processing of the human body. Science 2001; 293: 2470-2473. Downing PE, Wiggett AJ, Peelen MV. Functional magnetic resonance imaging investigation of overlapping lateral occipitotemporal activations using multi-voxel pattern analysis. J.Neurosci. 2007; 27: 226-233. Eimer M. Effects of face inversion on the structural encoding and recognition of faces. Evidence from event-related brain potentials. Brain Res.Cogn.Brain Res. 2000a; 10: 145-158. Eimer M. The face-specific N170 component reflects late stages in the structural encoding of faces. Neuroreport 2000b; 11: 2319-2324. Gauthier I, Tarr MJ, Moylan J, Skudlarski P, Gore JC, Anderson AW. The fusiform „face area“ is part of a network that processes faces at the individual level. J.Cogn.Neurosci. 2000; 12: 495-504. Haxby JV, Hoffman EA, Gobbini MI. The distributed human neural system for face perception. Trends Cogn. Sci. 2000; 4: 223-233. Herrmann MJ, Ehlis AC, Muehlberger A, Fallgatter AJ. Source localization of early stages of face processing. Brain Topogr. 2005; 18: 77-85. Itier RJ, Taylor MJ. Effects of repetition learning on upright, inverted and contrast-reversed face processing using ERPs. Neuroimage 2004a; 21: 1518-1532.

STUDIEN

UND


Itier RJ, Taylor MJ. Face recognition memory and configural processing: a developmental ERP study using upright, inverted, and contrast-reversed faces. J.Cogn.Neurosci. 2004b; 16: 487-502. Itier RJ, Taylor MJ. N170 or N1? Spatiotemporal differences between object and face processing using ERPs. Cereb.Cortex 2004c; 14: 132-142. Jacques C, Rossion B. Early electrophysiological responses to multiple face orientations correlate with individual discrimination performance in humans. Neuroimage 2007; 36: 863-876. Kanwisher N, McDermott J, Chun MM. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J.Neurosci. 1997; 17: 4302-4311. Kanwisher N, Tong F, Nakayama K. The effect of face inversion on the human fusiform face area. Cognition 1998; 68: B1-11. Kanwisher N, Yovel G. The fusiform face area: a cortical region specialized for the perception of faces. Philos.Trans.R.Soc.Lond. B Biol.Sci. 2006; 361: 2109-2128. Kovacs G, Zimmer M, Banko E, Harza I, Antal A, Vidnyanszky Z. Electrophysiological correlates of visual adaptation to faces and body parts in humans. Cereb.Cortex 2006; 16: 742-753. Leder H, Bruce V. When inverted faces are recognized: the role of configural information in face recognition. Q.J.Exp.Psychol.A 2000; 53: 513-536. Mangun GR. Neural mechanisms of visual selective attention. Psychophysiology 1995; 32: 4-18. Maurer D, Grand RL, Mondloch CJ. The many faces of configural processing. Trends Cogn.Sci. 2002; 6: 255-260. Minnebusch DA, Suchan B, Daum I. Losing your head: behavioral and electrophysiological effects of body inversion. J.Cogn.Neurosci. 2008. Minnebusch DA, Suchan B, Ramon M, Daum I. Event-related potentials reflect heterogeneity of developmental prosopagnosia. Eur.J.Neurosci. 2007; 25: 2234-2247. Moro V, Urgesi C, Pernigo S, Lanteri P, Pazzaglia M, Aglioti SM. The neural basis of body form and body action agnosia. Neuron 2008; 60: 235-246.

STUDIEN

UND


Morris JP, Pelphrey KA, McCarthy G. Occipitotemporal activation evoked by the perception of human bodies is modulated by the presence or absence of the face. Neuropsychologia 2006; 44: 1919-1927. Peelen MV, Downing PE. Selectivity for the human body in the fusiform gyrus. J.Neurophysiol. 2005; 93: 603-608. Peelen MV, Downing PE. The neural basis of visual body perception. Nat.Rev.Neurosci. 2007; 8: 636-648. Pitcher D, Charles L, Devlin JT, Walsh V, Duchaine B. Triple dissociation of faces, bodies, and objects in extrastriate cortex. Curr.Biol. 2009; 19: 319-324. Reed CL, Stone VE, Bozova S, Tanaka J. The body-inversion effect. Psychol.Sci. 2003; 14: 302-308. Reed CL, Stone VE, Grubb JD, McGoldrick JE. Turning configural processing upside down: part and whole body postures. J.Exp.Psychol.Hum.Percept.Perform. 2006; 32: 73-87. Righart R, de Gelder B. Impaired face and body perception in developmental prosopagnosia. Proc.Natl. Acad.Sci.U.S.A 2007; 104: 17234-17238. Rossion B, Delvenne JF, Debatisse D et al. Spatio-temporal localization of the face inversion effect: an event-related potentials study. Biol.Psychol. 1999; 50: 173-189. Rossion B, Gauthier I. How does the brain process upright and inverted faces? Behav.Cogn. Neurosci. Rev. 2002; 1: 63-75. Rossion B, Gauthier I, Tarr MJ et al. The N170 occipito-temporal component is delayed and enhanced to inverted faces but not to inverted objects: an electrophysiological account of face-specific processes in the human brain. Neuroreport 2000; 11: 69-74. Schwarzlose RF, Baker CI, Kanwisher N. Separate face and body selectivity on the fusiform gyrus. J.Neurosci. 2005; 25: 11055-11059. Stekelenburg JJ, de Gelder B. The neural correlates of perceiving human bodies: an ERP study on the body-inversion effect. Neuroreport 2004; 15: 777-780. Tanaka JW, Farah MJ. Parts and wholes in face recognition. Q.J.Exp.Psychol.A 1993; 46: 225-245.

STUDIEN

UND


Tarkiainen A, Cornelissen PL, Salmelin R. Dynamics of visual feature analysis and object-level processing in face versus letter-string perception. Brain 2002; 125: 1125-1136. Taylor JC, Wiggett AJ, Downing PE. fMRI analysis of body and body part representations in the extrastriate and fusiform body areas. J.Neurophysiol. 2007. Taylor MJ, Batty M, Itier RJ. The faces of development: a review of early face processing over childhood. J.Cogn. Neurosci. 2004; 16: 1426-1442. Thierry G, Pegna AJ, Dodds C, Roberts M, Basan S, Downing P. An event-related potential component sensitive to images of the human body. Neuroimage 2006; 32: 871-879. Tsao DY, Livingstone MS. Mechanisms of face perception. Annu.Rev.Neurosci. 2008; 31: 411-437. Varga AC. Declaration of Helsinki (Adopted by the 18th World Medical Assembly in Helsinki, Finland, and revised by the 29th World Medical Assembly in Tokyo, 1975). In: The Main Issue in Bioethics, revised edition. Paulist Press, New York, 1984. Yin RK. Looking at upside-down faces. J. Exp.Psychol. 1969; 81: 141-145. Yin RK. Face recognition by brain-injured patients: a dissociable ability? Neuropsychologia 1970; 8: 395-402. Yovel G, Kanwisher N. The neural basis of the behavioral face-inversion effect. Curr.Biol. 2005; 15: 2256-2262.

STUDIEN

UND


Studie 4

are human body forms processed holistically? evidence from the body coposite effect Denise A. Minnebusch und Irene Daum (2009a) Cortex, in Begutachtung

STUDIEN

UND


Are human body forms processed holistically? Evidence from the body composite effect Denise A. Minnebusch und Irene Daum, Cortex, In Begutachtung


STUDIEN

UND


ABSTRACT The present study aimed to investigate whether body forms are processed holistically. Evidence for holistic face processing comes predominately from the face composite effect: two identical top halves of a face are perceived as being different if they are presented with different bottom parts. This effect disappears if both parts are laterally shifted (misaligned) or if the stimulus is rotated by 180°. We investigated whether comparable composite effects exist for body forms using human bodies with and without heads. Recognition of upright bodies with heads was faster for misaligned compared to aligned presentations; this effect disappeared when the bodies were inverted. Composite effects did not emerge for bodies without heads. These findings provide evidence for holistic processing of naturalistic presentations of human body forms. It remains to be clarified, whether these observations are linked to holistic body processing or whether bodies with heads activate both face and body processing mechanisms.

STUDIEN

UND


INTRODUCTION Human bodies and faces are social cues which provide important information about the identity of other people, their age, gender, affect and intentions. It is generally accepted that faces undergo configural processing, and according to Maurer et al. (2002), three processing mechanisms can be distinguished. All faces share the same basic configuration or “first-order relations” (e.g., two eyes above the nose, the nose above the mouth). First-order relational information is sufficient for the recognition of common objects (e.g., houses, shoes, cars), but not for face or body shape recognition. The second mechanism has been termed “holistic processing” and refers to faces in terms of unique representations as opposed to a combination of single features. Finally, recognition of individual faces requires the encoding of spatial distances between internal features and minor variations in the shape of facial features. This configural processing mechanism is based on “second-order relations” and recruited for face identification. Evidence for configural face processing is mainly based on the face inversion effect (Yin, 1969; 1970): Upside down (inverted) presentation of stimuli disproportionally disrupts the recognition of faces compared to other objects. Inversion effects have recently also been observed for human body forms (Minnebusch et al., 2008; Reed et al., 2003; 2006). Recognition of inverted human bodies is slower and more difficult than the recognition of uprightly presented bodies (Reed et al., 2003; 2006). Since human body shapes are normally perceived with heads, body processing typically recruits both body and face processing mechanisms. Human bodies without heads (e.g., Downing et al., 2006; Taylor et al., 2007), on the other hand, are unnatural stimuli with unknown processing components. Body inversion effects were recently replicated for body forms with masked faces, while bodies without heads led to inverse effects, i.e., better recognition for inverted relative to upright stimuli, and the behavioral findings for the inversion effects were mirrored in the amplitudes of a body-sensitive ERP component (Minnebusch et al., 2008). These findings suggest that human body forms might be processed by specialized cortical mechanisms which differ at least partly from the mechanisms involved in face and object processing. Face inversion affects each of the three types of configural processing (Maurer et al., 2002), but it is as yet unclear which processing types are involved in the processing of body forms and whether they differ from face processing. Among the numerous configural face processing studies (Collishaw and Hole, 2000; Reed et al., 2003; 2006), holistic processing was mainly assessed on the basis of the face composite effect (Young et al., 1987): Two identical top halves of a face are perceived as being different if they are combined with different bottom parts. Since both stimuli are encoded holistically, recognition of individual facial parts is difficult, and the effect demonstrates that facial features cannot be processed independently. However, when the bottom and top halves are spatially shifted (i.e., misaligned), inter-

STUDIEN

UND


ference by holistic processing is reduced and the reported effect disappears. There is as yet no evidence of a comparable effect in the processing of human body shape. In summary, human bodies and faces might share the same recognition stream during early processing steps, in which perception involves the processing of first-order relations. Differences might, however, emerge in later processing steps, in which recognition is based on holistic processing and/or secondorder relations. The present study aims to investigate potential holistic processing of human body forms using a composite task. The effect of facial cues was explored by presenting human bodies both with masked faces and without heads.

METHOD Subjects Thirty-eight right-handed healthy young subjects (20 female; mean age 28.1, SD = 6.5) participated in this study. They were recruited by advertisements and received course credit for reimbursement. There was no history of neurological or psychiatric disorders in any of the subjects and all subjects had normal or corrected-to-normal vision. The study was performed in accordance with ethical standards laid down in the Declaration of Helsinki (Varga, 1975). Written informed consent was obtained from all subjects and the study was approved by the Ethics Committee of the Medical Faculty of the Ruhr-University, Bochum, Germany.

Stimuli Gray-scaled front view photographs of human bodies wearing dark clothes were presented on a white background. Each stimulus category consisted of 60 pictures at a size of approximately 3° x 3° of visual angle per item. Photographs were taken from a set of photographs of whole bodies, which was devised in our department. To minimize face processing during presentation of bodies, the faces on the pictures were masked or – in line with previous investigations (Downing et al., 2006; Minnebusch et al., 2008; Peelen and Downing, 2005; Taylor et al., 2007) – body shapes were presented without heads.

STUDIEN

UND


Figure 1: Illustration of the stimuli (top) and experimental design (down).

STUDIEN

UND


Human bodies with and without heads were displayed in an aligned and a misaligned version. Based on previous studies investigating the face composite effect (e.g., Rossion and Boremanse, 2008), an aligned version of each body picture was created by separating the upper and lower body parts by a small gap (1.8 mm). Similar to the stimuli used to assess the face composite effect (e.g., Rossion and Boremanse, 2008), the lower body part was shifted to the right in the misaligned version, starting at the middle of the upper body part (see Figure 1). The aligned and misaligned stimuli were used as the first stimulus in a delayed matching task. The top part of each body picture was then paired with a lower body part of another person of the same gender to create combinations of new stimuli. Human bodies with and without heads (aligned and misaligned) were presented uprightly and inverted, leading to eight different body representations (see Figure 1).

Procedure Subjects were seated in a sound-attenuating room facing a computer monitor at a distance of 80 cm. They were instructed to fixate the center of the screen. In each trial, pairs of body pictures were presented sequentially in random order. Each trial began with the presentation of a fixation cross, which was placed in the center of the screen for 100 ms (see Figure 1). It was followed by the first body picture which was presented for 400 ms and a scrambled mask for 200 ms. After an interval of about 550 ms (400 to 700), the second body stimulus was shown for 400 ms, followed by a an inter-trial interval of 1000 ms in which a black exclamation mark was presented in the center of the screen. The first and second stimuli always belonged to the same category (human body with or without head), showed subjects of the same gender and were presented in the same orientation (upright or inverted). In addition, both stimuli were either aligned or misaligned. The subjects’ task was to decide as fast and accurately as possible via key press, whether the top part of the first and second pictures were identical (same condition) or different (different condition), ignoring the lower body part. The bottom parts of the first and second stimuli were different in both conditions (same and different), whereas the top parts differed on half of the trials. Each first stimulus was used once in the same and once in the different condition. In total, two blocks of 240 trials (60 trials for each combination, see Figure 1) were administered.

STUDIEN

UND


Data analyses The percentage of correct responses and medians of reaction times (RTs) were assessed for each condition and each subject. In line with previous studies (Jacques and Rossion, 2007; Minnebusch et al., 2008), efficiency scores (RT divided by the proportion of correct responses) were calculated to obtain a single performance measure. A low efficiency score indicates good performance. The proportion of correct responses, RTs and efficiency scores were submitted to 2x2x2 repeated measures ANOVAs with Greenhouse-Geisser corrections with factors Stimulus Category (human bodies with heads vs. human bodies without head), Alignment (aligned vs. misaligned) and Orientation (upright vs. inverted).

RESULTS Percentage of correct responses The percentage of correct responses, RTs and efficiency scores are illustrated in Figure 2. ANOVA yielded a main effect of Orientation (F1,37 = 81.3, p < 0.001) as well as a significant interaction between Stimulus Category and Orientation (F 1,37 = 72.0, p < 0.001). Paired comparisons yielded better performance for uprightly presented compared to inverted human bodies, with larger effects for human bodies with heads (F1,37 = 120.1, p < 0.001) compared to human bodies without heads (F1,37 = 14.8, p < 0.001; see Figure 2). None of the other comparisons reached significance.

STUDIEN

UND


Figure 2: Percentage of correct responses and RTs (in ms) for human bodies with and without heads.

Reaction times ANOVA yielded main effects of Stimulus Category (F1,37 = 7.8, p = 0.008), Alignment (F1,37 = 5.5, p = 0.03) and Orientation (F1,37 = 35.9, p < 0.001). The Stimulus Category x Orientation interaction (F1,37 = 6.1, p = 0.02) and the interaction between Stimulus Category, Alignment and Orientation (F1,37 = 5.1, p = 0.03) were also significant. RTs were faster for human bodies with heads compared to human bodies without heads, for

STUDIEN

UND


misaligned bodies compared to aligned bodies, as well as for upright compared to inverted human body forms. To resolve the three-way interaction, ANOVAs were performed separately for human bodies with and without heads. For human bodies without heads, only the main effect of Orientation (F1,37 = 8.1, p = 0.007) reached significance, reflecting faster RTs for upright compared to inverted human bodies. For human bodies with heads, ANOVA yielded a significant Alignment x Orientation interaction (F1,37 = 4.1, p = 0.05), with faster RTs for upright compared to inverted human bodies and slower RTs for aligned compared to misaligned human bodies. Subsequent pairwise comparisons showed that the Alignment effect emerged for uprightly presented human bodies (t1,37 = 13.0, p = 0.01) but not for inverted human bodies.

Efficiency scores Analysis of efficiency scores yielded main effects of Stimulus Category (F1,37 = 5.0, p = 0.03) and Orientation (F1,37 = 5.4, p = 0.03) as well as an interaction between Stimulus Category and Orientation (F1,37 = 4.4, p = 0.04). Pairwise comparisons revealed a main Orientation effect, with better performance for upright compared to inverted human bodies with heads (F1,37 = 11.5, p = 0.002). None of the other comparisons reached significance.

DISCUSSION The present study aimed to investigate composite effects in body processing as potential evidence for holistic processing of human body shapes. Previous studies have indicated similarities in the early processing of human faces and bodies, while it is unclear whether such parallels would also apply to later processing steps involving holistic processing (Minnebusch et al., 2008; Peelen and Downing, 2007; Reed et al., 2003; 2006). In the face literature, holistic processing is defined as the phenomenon that single facial features cannot be encoded in isolation (Maurer et al., 2002). This view is mainly based on the face composite effect, i.e., the perception of two identical top halves of a face as being different when presented with different bottom halves (Young et al., 1987). This effect disappears, if the bottom parts are laterally shifted or if the faces are inverted (Goffaux and Rossion, 2006; Young et al., 1987). In the current investigation, we focused on the composite effect to investigate holistic processing of human bodies presented with and without heads. If human body forms, like human faces, are processed as unique representations, we would expect faster and more accurate processing of misaligned compared

STUDIEN

UND


to aligned body representations. The findings did not provide evidence of composite effects neither for the percentage of correct responses nor the efficiency scores for both categories (human bodies with and without heads; see Figure 2). However, for human bodies with heads, RTs were faster for the misaligned compared to the aligned stimuli, and this effect did not emerge for inverted body forms. This pattern shows clear similarities to previously reported face composite effects (Rossion and Boremanse, 2008), and offers some support to the assumption of holistic processing of human body shapes (presented with heads). Since a composite effect was found only for processing speed (RTs) and not for accuracy, holistic processing might be less important for body perception compared to face perception. There was no evidence of a composite effect for human bodies without heads and thus no evidence of holistic processing. Bodies without heads are unnatural and might activate different mechanisms compared to natural body stimuli (Minnebusch et al., 2008). Performance was generally faster and better for human bodies with heads compared to human bodies without heads, suggesting that the head might serve as a relevant cue for body shape identification, although the faces of our stimuli were masked and thus did not involve distinct facial features. The current pattern of findings offers some support to the idea that the processing mechanisms underlying the recognition of human body forms differ at least in part from those involved in face recognition. This is consistent with neuroimaging evidence of brain areas specialized for the processing of human bodies (e.g., Downing et al.,, 2001; Schwarzlose et al., 2005; Taylor et al., 2007), namely the extrastriate body area (EBA) and the fusiform body area (FBA). These areas can be distinguished from the fusiform face area (FFA) and the occipital face area (OFA) which have been associated with face perception (Kanwisher et al., 1997; Kanwisher and Yovel, 2006; Schwarzlose et al., 2005). Despite the differences in composite effects with respect to the accuracy and efficiency data, there are also similarities between the face and body perception data. As for faces, the responses associated with body processing were faster and more accurate for uprightly presented body forms compared to inverted body forms (see Figure 2). Since the inversion effect is thought to indicate configural processing, the present data offer further evidence of configural processing of body forms (with heads) similar to human faces (Minnebusch et al., 2008; Reed et al., 2003; 2006). The fact that our previous findings of faster and better processing of inverted compared to upright human bodies without heads (Minnebusch et al., 2008) could not be replicated in the current study might be due to procedural differences between the two studies (e.g., task differences, use of clothes as cues for identification in the task of the former study). The present results support the assumption that holistic processing is relevant for the perception of natural body stimuli (bodies with heads) though presumably not as critical as for the processing of

STUDIEN

UND


human faces. Faces and body shapes might share early processing mechanisms (including first-order relational information, structural information and - in part - holistic processing). It remains to be clarified whether the presence of the head activates face specific mechanisms during body processing and which mechanisms underlie the processing of the rather unnatural stimuli representing human bodies without heads.

STUDIEN

UND


REFERENCES Collishaw SM, and Hole GJ. Featural and configurational processes in the recognition of faces of different familiarity. Perception, 29: 893-909, 2000. Downing PE, Chan AW, Peelen MV, Dodds CM, and Kanwisher N. Domain specificity in visual cortex. Cerebral Cortex, 16: 1453-1461, 2006. Downing PE, Jiang Y, Shuman M, and Kanwisher, N. A cortical area selective for visual processing of the human body. Science, 293: 2470-2473, 2001. Downing PE, Peelen MV, Wiggett AJ, and Tew BD. The role of the extrastriate body area in action perception. Social Neuroscience, 1: 52-62, 2006. Goffaux V and Rossion B. Faces are “spatial” - holistic face perception is supported by low spatial frequencies. Journal of Experimental Psychology. Human Perception and Performance, 32: 1023-1039, 2006. Jacques C and Rossion B. Early electrophysiological responses to multiple face orientations correlate with individual discrimination performance in humans. Neuroimage, 36: 863-876, 2007. Kanwisher N McDermott J, and Chun MM. The fusiform face area: a module in human extrastriate cortex specialized for face perception. Journal of Neuroscience, 17: 4302-4311, 1997. Kanwisher N, and Yovel G. The fusiform face area: a cortical region specialized for the perception of faces. Philosophical Transactions of the Royal Society of London. Series B, Biological Science, 361: 2109-2128, 2006. Maurer D, Grand RL, and Mondloch CJ. The many faces of configural processing. Trends in Cognitive Science, 6: 255-260, 2002. Minnebusch DA, Suchan B, and Daum I. (2008). Losing your head: behavioral and electrophysiological effects of body inversion. Journal of Cognitive Neuroscience (in press). Peelen MV, and Downing PE. Selectivity for the human body in the fusiform gyrus. Journal of Neurophysiology, 93: 603-608, 2005.

STUDIEN

UND


Peelen MV, and Downing, PE. The neural basis of visual body perception. Nature Reviews. Neuroscience, 8: 636-648, 2007. Reed CL, Stone VE, Bozova S, and Tanaka J. The body-inversion effect. Psychological Science, 14: 302-308, 2003. Reed CL, Stone VE, Grubb JD, and McGoldrick, JE. Turning configural processing upside down: part and whole body postures. Journal of Experimental Psychology, Human Perception and Performance, 32: 73-87, 2006. Rossion B and Boremanse A. Nonlinear relationship between holistic processing of individual faces and picture-plane rotation: evidence from the face composite illusion. Journal of Vision, 8: 3-13, 2008. Schwarzlose RF, Baker CI, and Kanwisher N. Separate face and body selectivity on the fusiform gyrus. Journal of Neuroscience, 25: 11055-11059, 2005. Taylor JC, Wiggett AJ, and Downing PE. fMRI analysis of body and body part representations in the extrastriate and fusiform body areas. Journal of Neurophysiology, 98: 1626- 1633, 2007. Varga AC. Declaration of Helsinki (Adopted by the 18th World Medical Assembly in Helsinki, Finland, and revised by the 29th World Medical Assembly in Tokyo, 1975). In: The Main Issue in Bioethics, revised edition. Paulist Press, New York, 1984. Yin RK. Looking at upside-down faces. Journal of Experimental Psychology, 81: 141-145, 1969. Yin RK. Face recognition by brain-injured patients: a dissociable ability? Neuropsychologia, 8: 395-402, 1970. Young AW, Hellawell D, and Hay DC. Configurational information in face perception. Perception, 16: 747-759, 1987.

STUDIEN

UND


Studie 5

Neuropsychological mechanisms of visual face and body perception Denise A. Minnebusch und Irene Daum (2009b) Neuroscience and Biobehavioral Reviews, in Begutachtung

STUDIEN

UND


Neuropsychological mechanisms of visual face and body perception Denise A. Minnebusch and Irene Daum, Neuroscience and Biobehavioral Reviews, in Begutachtung


Keywords: body perception, configural processing, extrastriate body area, fusiform body area, inversion effect

STUDIEN

UND


Abstract Human faces and bodies provide important social cues, which contribute to the identification of other people, their age and gender as well as their intentions and affective states. The underlying neuropsychological mechanisms of face processing have been studied extensively and recent interest has also focused on the study of body shape perception. The present article aims to summarize and to critically evaluate the evidence for and against the specificity of body shape processing. Cognitive mechanisms, neurocognitive models and neuronal correlates of body processing will be compared with corresponding evidence related to human face processing. Clinical phenomena related to body shape perception will also be addressed. The available data base documents a range of similarities and differences between face and body perception with respect to the cognitive mechanisms, neuronal correlates and neuropsychological impairment patterns. The lack of a selective deficit in body perception is the most important difference between both categories. The sparse data base for human body shape perception does not yet allow any firm conclusions with respect to its underlying neuropsychological mechanisms.

STUDIEN

UND


Contents 1. 1.1 1.2

2. 2.1 2.1

Introduction Aims of the review Views on the cortical mechanisms of visual face and body perception

Brain correlates of face and body perception Evidence from neurophysiology Evidence from functional neuroimaging

3. Body perception and related neuropsychological processes 3.1 3.2 3.3

4. 4.1 4.2

Perception of biological motions Body perception and emotion Clinical studies

Integrative models, summary and open issues Model of face and body perception General summary and open issues

STUDIEN

UND


1.

Introduction

1.1

Aims of the review

Although human faces and body shapes are clearly very different visually, they also have certain features in common. Both faces and bodies provide multiple cues which are relevant for social communication and interaction. They both contribute to the recognition of other people and the identification of their identity, age, gender, intentions and emotional state. For both faces and body shapes, the differences between exemplars are metric rather than qualitative. Like faces, human bodies are symmetric and made up by critical elements, and the visual discrimination between different people requires finegrained distinctions based on the size and shape of these parts. The published neurocognitive research has so far predominantly focussed on the processing of faces (for reviews, see Kanwisher and Yovel, 2006; Rossion and Gauthier, 2002). In recent years, research efforts have been initiated to address the neuronal mechanisms underlying the processing of human body shapes (Peelen and Downing, 2007a). Both stimulus classes appear to be processed in specific cortical areas by specialized cognitive mechanisms (Kanwisher and Yovel, 2006; Peelen and Downing, 2007a). It is now generally accepted that face recognition depends on the detailed analysis of the configuration of facial features, i.e., on configural processing (Leder and Bruce, 2000; Maurer et al., 2002). It is as yet unclear whether the perception of human body forms is also based on configural processing mechanisms. This view is supported by converging findings from studies based on a range of different cognitive neuroscience methods (behavioral studies, electrophysiology, functional neuroimaging). A major aim of this review is the critical evaluation of the evidence for and against the idea of unique processing mechanisms for human body shapes and their potential links to configural face processing mechanisms. Each section of this review will initially outline the main findings from the face processing literature, followed by a more detailed review and analysis of the available literature on body shape processing. In the first section, relevant terms and definitions will be introduced and possible implications for the underlying mechanisms of face and body processing will be given by pointing out similarities and differences between both categories. The second section focuses on the neuronal correlates of face and body shape processing. In the third section, cognitive processes contributing to body shape perception and its dysfunction are addressed. The last section focuses on models of face and body shape processing and outlines open issues which require further research.

STUDIEN

UND


1.2

Views on the cortical mechanisms of visual face and body perception

The following section addresses the cortical mechanisms critically involved in face and body processing. The relevant terms and definitions will be outlined and research strategies for the study of the proposed mechanisms will be described. The terms configural and holistic processing which have been used synonymously in some studies, will be used following the definitions by Maurer et al. (2002), who consider holistic processing a sub-process of configural processing (see below).

a)

Face perception

The term configural processing refers to any process which involves the perception of relations among the features of a stimulus. It is contrasted with analytic processing, which is characterized by featurebased processing mechanisms (Leder and Bruce, 2000; Maurer et al., 2002). Face recognition might entail both analytical and configural processing (Collishaw and Hole, 2000; Tanaka and Farah, 1993). Maurer et al. (2002) have distinguished between three different types of configural face processing, located on a continuum (see Figure 1a): All faces share the same basic configurations, i.e., the same first-order relations, which specify the stimulus as a face. Processing first-order relations refers to the detection of the position of features in space which define a face (e.g., two eyes above the nose). Holistic processing refers to the fact that features of a face are bound together into a gestalt. In other words, facial features are integrated into an individual representation of the face as a whole. Holistic processing is based not only on internal features but also on external contours. Second-order relations refer to the spatial distances among internal features (e.g., distance between both eyes). Recognition of individual faces requires the encoding of information about minor variations in the shape or the spacing of features. In sum, face detection depends on the processing of first-order relations, whereas face identification requires the processing of second-order relations. However, it is still a matter of debate, whether the three types of configural processing are located on a continuum or whether they operate in parallel. It is possible that faces and bodies use similar but not identical mechanisms. Further studies are needed to clarify the exact location of these configural processing types. Another point on the configural processing continuum which was not considered by Maurer et al. (2002) refers to the structural information or hierarchical structure of an object (Reed et al., 2006). The authors described a configural processing continuum with the two extremes part-based processing and holistic processing. Mechanisms for body perception are located between these two extremes (see Figure 1b).

STUDIEN

UND


Face recognition depends not only on the fact that the mouth is located below the nose (first-order relational information), but also on the fact that the mouth is located in a particular position relative to the overall face structure. In addition, body parts are always attached to the same part of the body which defines its hierarchical structure, although the precise position of legs and arms may differ depending on posture or movement (Reed et al., 2006). First-order relationships provide information about the hierarchical structure of an object, whereas structural information defines the position of parts in the overall hierarchy. Structural information thus constitutes another point on the configural processing continuum between first-order relational information and holistic processing. Structural information - defined as information about the organization of object parts in relation to the overall object form and the spatial relation of each part relative to the other parts - plays an important role in face and body perception.

Figure 1: Illustration of the configural face and body processing continuum.

The face inversion effect is considered as critical evidence for configural face processing (Yin, 1969; 1970): Recognition of faces which are presented upside down is significantly more difficult than recognition of upright faces, i.e., subjects need longer to process inverted faces and they make more errors in face

STUDIEN

UND


matching tasks or similar procedures. Although most objects are more difficult to recognize if inverted compared to upright presentation, inversion disproportionally disrupts the recognition of faces relative to other objects. Inversion effects are thought to reflect configural processing and they clearly affect each of the different configural processing mechanisms (Maurer et al., 2002; Reed et al., 2006). The inversion effect is the most robust phenomenon in the face processing literature; it is observed with old-new recognition tasks and two-alternative forced-choice tasks with and without delay and for both familiar and unknown faces (for review, see Rossion and Gauthier, 2002). There is comprehensive evidence for the contribution of all subtypes of configural processing to human face perception (see Maurer et al., 2002). Reed et al. (2006) demonstrated the relevance of structural information (see 1.2b). Evidence for holistic face perception stems from studies of the face composite effect (Young et al., 1987): Two identical top halves of a face are perceived as being different if they are combined with different bottom parts. This effect indicates that facial features cannot be processed independently. If the bottom and top halves are laterally shifted, this effect disappears. An alternative hypothesis (“expertise hypothesis”) suggests that faces are special because people have more experience with face processing compared to processing of members of other stimulus classes (Diamond and Carey, 1986). Whether or not expertise underlies human face processing is a matter of an ongoing debate which has been the subject of a recent review (see Gauthier and Bukach, 2007; McKone et al., 2007). Both faces and bodies provide multiple cues which are necessary for social interactions and interpersonal relationships. In accordance to face perception, people might also be experts for human bodies, since we have more experience with body perception than with the processing of other objects. To our knowledge, the expertise hypothesis has so far not been discussed in relation to human body form perception and it will therefore not be considered further in this review.

b)

Body perception

Recent evidence suggests that configural processing is critical not only for faces, but also for the recognition of human body forms (Reed et al., 2003; Stekelenburg and de Gelder, 2004). Reed et al. (2006) showed that faces and bodies partly share the same recognition processing stream. Configural recognition of body postures is dependent upon the structural hierarchy of body parts, but further evidence is needed to support this assumption (see Figure 1). Configural processing of body forms has also been supported by studies of inversion effects which yielded similar effects for faces and bodies. Reed et al. (2003; 2006) reported slower reaction times and

STUDIEN

UND


higher error rates for decisions about inverted compared to upright human body positions. The inversion effect per se does not allow any firm conclusions concerning the amount of information necessary for configural processing, although the whole body form appears to be necessary, since inversion effects were not observed for isolated body parts or scrambled bodies (Reed et al., 2006). Further analyses stressed the importance of structural information and the processing of first-order relations between body parts and the whole body context. Configural body processing thus seems to rely on the structural hierarchy of body parts, not on the isolated parts themselves. To our knowledge, there are as yet no studies of holistic body processing and second-order relations for body processing. The available findings offer some support for the idea that human bodies - like faces - are processed configurally. During early processing stages, human bodies and faces might share the same recognition stream (first-order relational information and structural information) while dissociations might occur at later processing stages (Figure 1). Summary: Body and face perception appear to involve both analytical and configural processing mechanisms. Although bodies and faces produce comparable inversion effects, they may be processed by different configural processing mechanisms. It should be noted that the reported body inversion effect is much smaller than the face inversion effect (Reed et al., 2003). Further studies are needed to clarify which type of configural processing applies to human body recognition.

2.

Brain correlates of face and body perception

Investigations using event-related potentials (ERPs) and functional magnetic resonance imaging (fMRI) have made important contributions to our knowledge about the time course and brain correlates of the processes involved in face and body perception.

2.1

Evidence from neurophysiology

a)

Face perception

Human faces elicit a negative ERP peaking about 170 ms (termed N170) after stimulus onset, with maximum amplitudes in occipitotemporal areas (Bentin et al., 1999; Eimer, 2000a; b; Minnebusch et al., 2007).

STUDIEN

UND


The neural source of the N170 is unclear, as is its exact relationship to face-sensitive regions identified in fMRI. N170 latencies are longer for inverted relative to upright faces. This effect is not observed for objects (Bentin et al., 1996; Eimer, 2000a; Itier and Taylor, 2004a, b; Rossion et al., 2000b). Most but not all studies also reported larger N170 amplitudes for inverted relative to upright faces (Itier and Taylor, 2004a; b; Marzi and Viggiano, 2007; Rossion et al., 1999; 2000b). The N170 is generated by brain processes involved in late stages of structural encoding, where representations of global face configurations are generated for face recognition (Eimer, 2000c; Jacques and Rossion, 2007). Furthermore, perception of human faces also affects the P1, an early endogenous ERP component appearing around 100-120 ms after stimulus onset. The P1 is larger for faces compared to other stimuli and enhanced and delayed for inverted compared to upright faces (Itier and Taylor, 2004a; c; Rossion et al., 1999; 2000b). The P1 seems to be an early global response to faces and could thus reflect the perception of a stimulus as a face (Herrmann et al., 2005; Itier and Taylor, 2004c). Source localization yielded a source of the P1 in the fusiform gyrus, while the N170 is localized in a parieto-temporal-occipital network comprising the fusiform gyrus and the lateral occipito temporal cortex (Herrmann et al., 2005; Iidaka et al., 2006; Pizzagalli et al., 2002). The similarities between the P1 and N170 raise the question of the time window in which processing differences between upright and inverted faces begin to emerge, i.e., the time window of the onset of configural processing. Jacques and Rossion (2007) investigated the relationship between behavioral and electrophysiological inversion effects for different face orientations. Both the P1 and the N170 were affected by face orientation, but significant correlations of behavioral performance (reaction times, error rates and efficiency score based on both) and ERP amplitudes only emerged for the N170 window, suggesting the behavioral face inversion effect occurs during perceptual face encoding. Another ERP component associated with face perception is the N250 (Itier and Taylor, 2004a; Schweinberger et al., 2002; Tanaka et al., 2006) which indexes face familiarity (Schweinberger et al., 2002; Tanaka et al., 2006). There are as yet no studies of a comparable potential associated with body perception.

b)

Body perception

Thierry et al. (2006) described an ERP component evoked by stimuli representing the form of the human body without heads (termed N190). The location of the N190 differed from the source of the N170 in the right posterior extrastriate cortex, although the exact sites of both sources remain to be determined in detail using combined fMRI and ERP methodology. Similar to the N170 evoked by faces, the N190 generalizes to some degree to schematic descriptions of the human body form (Thierry et al., 2006). Furthermore, the latency of the P1 elicited by objects was significantly longer than the P1 evoked by human body forms and faces.

STUDIEN

UND


As concerns the underlying information processing mechanisms, the idea of configural processing of human body shapes was supported by larger N190 amplitudes and longer latencies for inverted compared to upright human bodies (Righart and de Gelder, 2007; Stekelenburg and de Gelder, 2004). Body inversion did not have a significant effect on P1 latency and amplitude (Righart and de Gelder, 2007). Both studies used photographs representing the whole body and faces were masked to minimize the involvement of face processing. However, fMRI studies indicated that contextual cues (such as masked faces) can elicit object-specific neuronal responses (Cox et al., 2004), and occipitotemporal activation evoked by body shapes is modulated by the presence or absence of the face (Morris et al., 2006). It is therefore possible that even masked faces might activate face-specific processing mechanisms to some degree, and it is unclear whether bodies, faces or both are critical for the reported effects. In addition, the naturalistic presentation of human body shapes includes heads, and body processing is therefore typically based on both body and face processing mechanisms. Human bodies without heads are thus unnatural stimuli, with different processes being evoked compared to natural body-with-head stimuli. A direct comparison of human bodies with and without heads yielded the typical behavioral and electrophysiological inversion effects for human bodies with masked faces (Minnebusch et al., 2008). For human bodies without heads, the inversion effect was inversed, with better recognition performance for inverted relative to upright bodies, presumably due to the disturbing nature of the unnaturalistic presentation. Taken together, the available findings imply that human body forms are processed by specialized cortical mechanisms which are at least partly separable from the mechanisms involved in face and object processing, but the precise pattern of similarities and differences remains to be determined.

2.2

Evidence from functional neuroimaging

a)

Face perception

Functional neuroimaging studies have identified a region in the fusiform gyrus, the fusiform face area (FFA), which is activated more strongly by passive viewing of faces compared to objects (Haxby et al., 2000; Kanwisher et al., 1997; Kanwisher and Yovel, 2006; see Figure 2). Bilateral FFA activation correlates with correct face identification, but not with correct object identification (Grill-Spector et al., 2004), with highest correlations in the right hemisphere (Haxby et al., 1999; Kanwisher et al., 1997; Rossion et al., 2000a). In addition to the FFA, the superior temporal sulcus (STS) and the inferior and middle occipital gyri show significant activations during face processing (Haxby et al., 1999; Kanwisher et al., 1997), also with a right hemispheric dominance. The face-selective inferior occipital area termed occipital face area (OFA; Gauthier et al., 2000) is sensitive to physical features of face stimuli (Figure 2), whereas the

STUDIEN

UND


STS appears to process dynamic aspects of facial information (e.g., expression, gaze). Although there is consensus that the FFA is selectively involved in face perception and identification (Grill-Spector et al., 2004; Kanwisher et al., 1997; Kanwisher and Yovel, 2006), its precise functional role remains to be fully explored. There is evidence that the FFA might be linked to view-dependent face representations (Pourtois et al., 2005). In an alternative view, Gauthier et al. (1999) suggested that FFA activation reflects expertise with processing of visual stimuli. The FFA might be activated by any stimulus which the subject perceives as distinct individual rather than as generic example of a category. Faces activate this region because all people are experts in face perception.

Figure 2: Illustration of the locations of the face sensitive OFA and FFA as well as the body sensitive EBA and FBA.

STUDIEN

UND


b)

Body perception

Similar to face perception, perception of human body forms has been associated with specific cortical processing units (Downing et al., 2001; 2006a; Peelen and Downing, 2005; 2007a; Schwarzlose et al., 2005). The extrastriate body area (EBA) is more strongly activated during perception of human bodies and body parts compared to objects or faces (Downing et al., 2001; Urgesi et al., 2004). The EBA seems to be a specialized system for the identification of a stimulus as a human body. EBA activations generalize to line drawings, stick figures and silhouettes, indicating that the EBA has an abstract body representation (Downing et al., 2001). EBA might thus be involved in the extraction of visual information about other individuals. Interestingly, EBA activation is modulated by the presence or absence of a face, with activation being stronger for bodies without heads compared to bodies with heads, presumably because attention is captured by a face when one is present, thereby reducing the activity of body selective regions (Morris et al., 2006). Kitada et al. (2008) identified two regions in the lateral occipitotemporal cortex which are sensitive to visual and haptic body perception (EBA and the haptic body region), whereas both areas differed in their response patterns. Comparable results are reported for faces and the fusiform gyrus. These finding indicates that face and body sensitive regions seem to be partly modality-independent. Anatomically, the EBA is located at the posterior end of the inferior temporal sulcus and overlaps with the motion selective area (MT) and with the object-form selective area (LO; Downing et al., 2007; Peelen and Downing, 2007a). Overlapping activation in this area might reflect a common neuronal system engaged in a common process. Alternatively, processing mechanisms linked to this region might differ between categories (Downing et al., 2007). However, studies using multi-voxel pattern analysis (MVPA) indicated functional independence in overlapping extrastriate cortical regions (Downing et al., 2007; Peelen et al., 2006; Peelen and Downing, 2007a; b). MVPA aims to decode the information that is represented in the subject’s brain at a particular point of time. Instead of focusing on individual voxels, MVPA uses patternclassification algorithms to determine the information that is represented in a given pattern of activity (Norman et al., 2006). Using this method, Downing et al. (2007) showed that the form of a human body, visual motion and object form elicited independent patterns of fMRI activation. Recently, a second area sensitive to human body form was identified, the fusiform body area (FBA; Peelen and Downing, 2005; Schwarzlose et al., 2005). The FBA is located in the posterior fusiform gyrus (Peelen et al., 2006; Peelen and Downing, 2005; Schwarzlose et al., 2005) and responds to the form of the whole body rather than body parts (Taylor et al., 2007). Faces and human body forms (stimuli presented without heads/faces) appear to be processed in adjacent and overlapping but distinct networks within the fusiform gyrus (Peelen and Downing, 2005; 2007a; Schwarzlose et al., 2005). In other words, faceand body-selective activations can be dissociated both anatomically and functionally.

STUDIEN

UND


Summary: Electrophysiological studies have indicated that the time course of face and body perception differs from the time course of object perception. For both stimulus classes, the P1 and N170/N190 are enhanced compared to ERPs evoked by other objects. Findings from functional neuroimaging have suggested specialized cortical networks for face and body perception. Taken together, the available evidence suggests that faces and human body forms are processed in adjacent and partly overlapping networks.

3. Body perception and related neuropsychological processes The following section will briefly address the neurocognitive mechanisms of biological motion perception and the mirror-neuron network, emotional body language and clinical phenomena related to body perception. Detailed reviews of these issues can be found elsewhere (Blake and Shiffrar, 2007; de Gelder, 2006; Rizzolatti and Craighero, 2004; Schindler et al., 2008; Vuilleumier and Pourtois, 2007).

3.1

Perception of biological motions

The movements of other people are of particular interest to humans. Although we can detect multiple information from static body postures, motion provides more reliable and more detailed information about intentions, emotions and actions of others. Investigations addressing these issues are typically based on point-light animations, where biological activity is presented by small light tokens placed on the major body parts of an actor (Johansson, 1973). Observers are able to recognize the identity and sex of a point-light walker or information derived for point-light faces (for a detailed review, see Blake and Shiffrar, 2007). Motion represented by a moving human body form made up by only a few moving point lights has been termed “biological motion” (Johansson, 1973; Lange et al., 2006). Similar to perception of static body postures, action perception is sensitive to stimulus inversion, supporting the assumption of configural body processing (Pavlova and Sokolov, 2000; Sumi, 1984). Inconsistent with this view, Jokisch et al. (2005) reported reduced N170 amplitudes for inverted compared to upright displays of walking stimuli. Along similar lines, Minnebusch et al. (2008) reported higher N170 amplitudes for upright human bodies without heads compared to inverted bodies without heads. Both types of stimulus material of both studies entail information about human body shape but not about the face, raising the question,

STUDIEN

UND


whether face processing at least partly underlies the observed body inversion effect in other studies (but see 2.1b). Troje and Westhoff (2006) reported inversion effects for normal and scrambled pointlight walkers, with observers having to extract information about the walking direction. In scrambled point-light displays, local motion is kept intact, whereas configural information is disrupted. These data might thus be interpreted as evidence against configural body processing. Since the inversion effects of point-light walkers seems to depend only on the motion of the feet and might be independent from the structure of the dot pattern, it is unclear whether these findings can be directly compared to other studies of body inversion (Troje and Westhoff, 2006). Viewing human body actions activates the STS, the parietal cortex and the ventrolateral premotor cortex in addition to the EBA and FBA. Specialized visual mechanisms are found in the STS which produce selective neuronal responses to moving images of faces and bodies (Puce and Perrett, 2003; Thompson et al., 2005). STS mechanisms generalize to biological motion of point light displays and static biological stimuli (Chan et al., 2004; Chao et al., 1999; Haxby et al., 1999; Kanwisher et al., 1997) The STS receives input from form areas (e.g., FFA and OFA, Fairhall and Ishai, 2007) and motion areas (Lange and Lappe, 2006; Thompson et al., 2005) and uses form cues – such as body configuration – to process biological motion (Thompson et al., 2005). Additionally, the STS shows stronger activations for biological motions compared to non-biological motions, suggesting that the STS is involved in the perception of face and body movements (Pelphyrey et al., 2003; Puce et al., 1998). However, it is also activated by static images of the face and body, indicating that it responds to stimuli which signal the actions of another individual (for review, see Allison et al., 2000). Although, both face and body movements activate the STS, a recent study reported distinct activation pattern for face and hand movements in this area (Thompson et al., 2007). The right mid-STS seems to be sensitive to face movements, whereas the right posterior inferior temporal sulcus showed higher activations for hand movements relative to face movements. The parietal cortex and the ventrolateral premotor cortex are involved in the perception and execution of action (Rizzolatti and Craighero, 2004). When individuals observe an action of another person, the motor cortex, more specifically the mirror-neuron system, becomes active, i.e., observation of an action activates parts of the same cortical network that is active during its execution. Mirror-neurons are visuomotor neurons, originally discovered in the premotor cortex of the monkey (area F5; for a detailed review, see Rizzolatti and Craighero, 2004). The human mirror-neuron system involves the rostral part of the inferior parietal lobule and the caudal part of the inferior frontal gyrus (Rizzolatti et al., 2001; Rizzolatti and Craighero, 2004). In humans, the system plays an important role in action understanding and action imitation, facilitating the understanding of the meaning of the actions and emotions of others (see 3.2) by direct simulation of the observed events without a need for conceptual reasoning. Simulation involves the association between the first- and third-person experience by linking personal actions and

STUDIEN

UND


emotions with actions and emotions of other people (Gallese et al., 2004). Action understanding entails the mapping of observed actions onto the observer’s motor representations. However, it is still unclear whether action perception and action execution activate the same neuronal population or whether they activate distinct but overlapping populations of perceptual and motor neurons (Dinstein et al., 2008). First evidence for human mirror neurons in the right inferior parietal lobe, which encode specific actions regardless of whether they are observed or executed, is reported by Chong et al. (2008). Further studies are needed to access the specific function of these neurons in humans. The EBA is not only activated by the perception of body parts, but also during observation of goal-directed movements of body parts (Astafiev et al., 2004). EBA activations are, however, also modulated by limb movements towards a visual stimulus, even in the absence of visual feedback from the movement. These findings suggest a possible influence of actions on visual cortex (Astafiev et al., 2004; but see Peelen and Downing, 2005) and a link between the EBA and the mirror-neuron network, although there are important processing differences between both systems with respect to action perception and self-generated motions. According to Urgesi et al. (2007a), the EBA processes the actor’s identity whereas the premotor cortex is critical for the discrimination of body actions. By representing the static structure of the human body form, the EBA plays a unique role in the perception of action (Downing et al., 2006b). It is thus not directly involved in the dynamic representation of biological actions, but codes the visual presence of the human form (Chan et al., 2004; Downing et al., 2006b). The EBA is involved in the distinction between self-generated and other-generated movements, thereby representing the human body in an integrative and dynamic manner (David et al., 2007).

3.2

Body perception and emotion

Visual areas for face and body perception are differentially involved in the perception of identity, movements and expressions of other people. The affective state of others is expressed in facial features, voice intonation as well as body posture and movement, i.e., emotional body language. A single system might be responsible for the perception and production of emotional body language (Gallese et al., 2004), and this system might be part of the mirror-neuron network. Perception of emotion is based on the same motor representation which is associated with emotion expression, similar to action understanding and execution (see 3.1). We thus use our own body for perception of other people and for understanding their emotions and intentions (see de Gelder, 2006). The mirror-neuron system plays an important role in the perception of facial expressions and body language (Gallese et al., 2004). Additional activations are observed in the amygdala, the insula, the fusiform gyrus, and the STS (Fairhall and Ishai, 2007; Gallese et

STUDIEN

UND


al., 2004; Grezes et al., 2007; Hadjikhani & de Gelder, 2003; Haxby et al., 2000; 2002; Ishai et al., 2005; Ishai, 2008), with more pronounced activations for dynamic compared to static facial or bodily expressions in movement sensitive areas like the STS and premotor cortex (Grezes et al., 2007). The insula appears to be particularly sensitive to the perception of disgust (for a recent review, see Vuilleumier and Pourtois, 2007), while the amygdala is disproportionately involved in the perception of fear and more generally in social cognition (Morris et al., 1996). A recent meta-analysis indicated that the amygdala responds to both positive and negative stimuli, with a preference for emotional faces (Sergerie et al., 2008). Enhanced activation in the FFA for emotional faces might be triggered through feedback connections from the amygdala (Vuilleumier and Pourtois, 2007). Detection of fear in faces and body expressions might be encoded at early stages of visual processing (Meeren et al., 2005; van Heijnsbergen et al., 2007). Schindler et al. (2008) proposed a biologically inspired, hierarchical model of the recognition of emotion expressed by body postures. According to this computational model, discriminations between neutral expressions and the six basic emotions (anger, disgust, fear, happiness, sadness and surprise) can be achieved by input based on static views of body postures, with emotional body language being based on low-level form processing. Additionally, de Gelder (2006) postulated a two system theory of emotional body language, consisting of two separate emotional circuits cooperating with each other: a rapid and non-conscious automated system including predominately subcortical structures (superior colliculus, pulvinar, striatum, basolateral amaygdala) and a cortical system (frontoparietal motor system and connectivity between the amygdala and the dorsolateral prefrontal and ventromedial prefrontal cortex) with reciprocal connections to the first system. The second system processes emotional body language in more detail and calculates behavioural consequences of an emotion by decoding emotional stimuli in combination with past experience and memory. So far, perception of emotion has mainly been discussed in the face and voice intonation literature (e.g., Dailey et al., 2002; Eimer and Holmes, 2002), with data on emotional body perception being sparse. Further studies are clearly needed to elucidate similarities and possible differences between emotional face and body perception.

3.3

Clinical studies

The detailed assessment of patients suffering from selective face and/or body processing would provide important insights into the underlying neurocognitive mechanisms. Until now, a selective deficit in body perception which is comparable to prosopagnosia has not been observed (Peelen and Downing, 2007a, but see Moro et al., 2008). Since body perception is presumably mediated by a dis-

STUDIEN

UND


tributed cortical network, damage to different components of the network may lead to distinct body shape processing deficits. Face processing deficits have been studied extensively in individuals suffering from prosopagnosia (Behrmann and Avidan, 2005; Duchaine and Nakayama, 2006; Kress and Daum, 2003). Prosopagnosia is a rare deficit in the ability to recognize a person by the face only (Bodamer, 1947). Acquired prosopagnosia is frequently linked to bilateral damage to occipito temporal areas. Face perception deficits of unknown etiology are referred to as developmental prosopagnosia (DP; Behrmann et al., 2005; Behrmann and Avidan, 2005), and DP has been linked to impairments of configural face processing. Given the closely related representations of faces and bodies (Peelen and Downing, 2007a; Schwarzlose et al., 2005), DP individuals may also have difficulties with configural body processing. This hypothesis has been supported by a recent study of altered ERPs evoked by both faces and bodies in DP subjects (Righart and de Gelder, 2007). Configural processing may thus be relevant for a range of visual stimuli in addition to faces. Faces and body shapes might both trigger configuration-based visual strategies which are critical at initial stages of stimulus encoding. Further studies will have to clarify whether the reported ERP changes are accompanied by behavioral changes, e.g., with respect to stimulus inversion effects, which would provide strong support for the configural body processing hypothesis. In a recent study of processing of emotional faces and body language in DP (van den Stock et al., 2008), neutral but not emotional face expressions elicited lower FFA activations of the DP group compared to the control group. Interestingly, bodies evoked higher activations in a face-sensitive area in the inferior occipital gyrus in the DP subjects, and neutral face expressions yielded activations in the body-sensitive EBA. Taken together, this pattern suggests that a reduced categorical segregation of face- and body-sensitive mechanisms in DP might be responsible for the observed difficulties in these subjects. A recent study by Moro et al. (2008) combined behavioral data with lesion mapping techniques. Lesions involving the EBA were associated with body form agnosia and lesions involving ventral premotor areas were associated with body action agnosia. Body form agnosia was assessed using a matching-to-sample task involving body parts, face parts and object parts as stimuli to investigate specific disorders in body and face part discrimination. Tasks based on whole bodies and faces might be necessary to clarify the precise deficits of these patients and to investigate deficits in configural processing. A direct comparison with prosopagnosia patients might be useful to determine whether body form agnosia can be dissociated from prosopagnosia. In a second experiment, the authors reported a double dissociation between body form and body action agnosia, providing strong support for the assumption of two different networks for action perception and body form perception (Urgesi et al., 2007). Lesions in the EBA appear to be associated with deficits in body form recognition and lesions in the ventral premotor cortex appear to be associated with deficits in action perception.

STUDIEN

UND


Face perception problems are frequently observed in autism. Autism is a neurodevelopmental disorder characterized by problems with social interaction, verbal and non-verbal communication and stereotypic behaviors (DSM-IV; American Psychiatric Association, 1994). There is evidence for atypical face and configural processing, as indicated by reduced face inversion effects (see Dawson et al., 2005) or an absence of face composite effects (Teunisse & de Gelder, 2003). Interestingly, the assessment of inversion effects indicates that configural body processing may be preserved in autism even if configural face processing is significantly impaired (Reed et al., 2007). These data further support the assumption of dissociable processing mechanisms for face and body perception, although the data base is as yet small. In line with the “expertise hypothesis” (see 1.2), it is conceivable that the frequent everyday exposure to human body forms leads to visual expertise for body processing. In other words, the body inversion effect might be linked to a mapping of the perceptual input onto a spatial representation which captures specific structural relations which are known from the observation of bodies of other people. These relations are disrupted when bodies are presented upside down. However, visual expertise with human bodies might be acquired not only from the frequent visual experience with other people’s bodies, but also from experiencing one’s own body, e.g., when using proprioceptive information about our own body when perceiving the bodies of others (see 3.1). In a detailed assessment of a patient suffering from deafferentiation, Bosbach and colleagues (2006) did, however, not find evidence for the hypothesis that static body posture recognition relied on covert proprioceptive inputs. Patients with Huntington’s disease show motor impairments together with cognitive and affective changes and deficits of recognition of emotions in faces and prosody (Campanella and Belin, 2007; Gobbini and Haxby, 2007; Vuilleumier and Pourtois, 2007). As outlined in section 3.2, human body postures provide multiple cues about the emotional state of other people and the mirror-neuron system is involved in the perception of body recognition and of emotional body language (Gallese et al., 2004). The combined deficit of body shape and emotional body language recognition and the significant correlations with motor deficits in patients with Huntington’s disease (de Gelder et al., 2008) indicated a strong link between mechanisms of emotional recognition and the motor system, although the anatomical link remains to be elucidated. Finally, the unitary experience between self and body can be disrupted and a person seems to see his/her body and the world from a location outside the physical body (Blanke et al., 2004; 2005). This phenomenon is termed out-of-body-experience. The temporoparietal junction might be an important structure for the self and impaired processes at this structure might lead to pathological disorders like out-of-bodyexperience (Blanke et al., 2005). Additionally, multisensory and cognitive processing of bodily information might be necessary for spatial unity (Lenggenhager et al., 2007).

STUDIEN

UND


Summary: Body movements provide important cues about the age and gender of an actor and of his/ her intentions, emotions and actions. Interestingly, action perception and action execution activate the same cortical network (mirror-neuron network). Furthermore, mirror mechanisms play an important role in the perception of one’s own emotions and the perception of emotions of others. Action and the understanding of emotions might involve the mapping of observed actions and emotions onto personal motor and emotional representations. Body perception activates a widely distributed cortical network, with each part of the network mediating a specific aspect of body processing, yielding complex neuropsychological patterns. Studies of patients with focal lesions are clearly needed to determine the functional-anatomical mappings in greater detail.

4.

Integrative models, summary and open issues

4.1

Model of face and body perception

a)

Face perception

Haxby et al. (2000) proposed a hierarchical model of face perception based on a core system and an extended system. The core system entails the inferior occipital gyrus (i.e., the OFA), the lateral fusiform gyrus (the FFA) and the STS. The extended system comes into play if additional information is extracted from faces (e.g., facial expression, lip-reading, or name and biographical information). The model assumes a distinction between the representation of invariant aspects of faces (predominantly the core system), which underlie the recognition of the identity of a person, and the representation of variable facial features, which play a more central role in social communication (Haxby et al., 2002). The core system is specialized for face perception whereas the extended system is multimodal and can therefore be accessed by faces and other stimuli (Fox et al., 2008; Gobbini and Haxby, 2007; Haxby et al., 2000). The authors suggested that both the fusiform gyrus and the STS receive input from the inferior occipital gyrus. This view was supported by a recent fMRI study (Fairhall and Ishai, 2007). An alternative hypothesis suggests that the right inferior occipital gyrus is reciprocally connected to the right fusiform gyrus (Rossion et al., 2003), with both regions being necessary for intact face perception. In addition, Rossion (2008) described a direct pathway from early visual areas to the fusiform gyrus. In general, face perception appears to activate a distributed cortical network in the human brain and the integrated activity of the various regions of this network is necessary for successful face perception (Fairhall and Ishai, 2007; Ishai, 2008).

STUDIEN

UND


b)

Body perception

Until now, two models of body perception have been outlined. The first model, suggested by Taylor et al. (2007), is based on similarities between face and body perception and focuses on the EBA and FBA. The second model proposes that visual processing of body shapes engages an extensive fronto-temporoparietal network (Urgesi et al., 2004; 2007a). Two cortical regions appear to be specialized for the perception of human bodies (FBA and EBA) and body parts (EBA; for review, see Peelen and Downing, 2007a). Taylor and colleagues (2007) have emphasized the parallels of face and body processing given the similarity of the functional neuroanatomy (see 3.2) with analogies between the face-sensitive OFA and the body-sensitive EBA and between the FFA and the FBA. When the amount of body information was varied gradually from single fingers to the whole body form, EBA activation remained comparable, while FBA showed a sharp increase when whole torsos were presented. The EBA thus seems to be involved in the basic analysis of body features (similar to the OFA in face processing), whereas the FBA (similar to the FFA) appears to be implicated in processing the configuration of body parts as a whole (Taylor et al., 2007). Functional connectivity between the EBA and FBA needs to be explored to determine the precise relationship between the regions in question, as has been analyzed for face-specific regions (Fairhall and Ishai, 2007). Urgesi et al. (2007) have suggested that the human brain comprises two dissociable and independent routes which are specialized for processing human bodies: One route processes the whole body in a configural manner and involves dorsal stream brain areas and the mapping of observed postures onto the observer’s sensorimotor representation. Configural processing of bodies thus depends upon sensorimotor representations rather than visual representations in the extratriate visual cortex, which are involved in local body processing. The second route is specialized for processing of local features of the body, such as body part and body form and is associated with the EBA. Evidence for this model are provided by patients with focal lesions in the EBA and ventral premotor cortex (Moro et al., 2008). Moro and colleagues (2008) reported a double dissociation between deficits in body form perception (associated with lesions including the EBA) and deficits in action perception (associated with lesions including the ventral premotor cortex) yielding important evidence for a dissociation of these two types of perception.

4.2

General summary and open issues

Human faces and human bodies provide multiple cues about other people. In the last decades, research efforts have mainly focussed on the cognitive mechanisms and neuronal correlates of face perception. It

STUDIEN

UND


is generally accepted that the human face represents a special stimulus which is processed by specialized cortical mechanisms. Recent research suggests that the perception of human bodies also seems to differ from the perception of other stimulus classes, but the pattern of similarities and differences between face and body processing remains to be fully determined. Face perception involves both analytical and configural processing, with the latter playing a more critical role. Maurer et al. (2002) described a configural face processing continuum consisting of three types of configural perception: first-order relations, holistic processing and second-order relations. The face inversion effect supports the assumption of configural face perception and affects each type of the configural processing continuum. A comparable inversion effect was also observed for the perception of human bodies indicating that configural processing is important for both face and body perception. Until now the impact of first-order relations on body perception has been addressed. However, it is yet unknown, whether holistic processing is important for configural body perception. Further studies are needed to determine whether configural body perception differs from configural face perception at later processing stages. Face perception activates specialized cortical regions (OFA and FFA) and elicits specific ERP components (primarily the P1 and N170). Recent EEG studies yielded evidence that configural face (e.g., Jacques and Rossion, 2007) and body processing (Righart and de Gelder, 2007; Stekelenburg and de Gelder, 2004) occur between 170ms and 190ms after stimulus onset. An inversed inversion effect was found for human bodies without heads (Minnebusch et al., 2008). It is a matter of debate, whether the reported electrophysiological body inversion effect might rely on face processing mechanisms. Human bodies without heads represent an unnatural visual stimulus and may thus evoke effects unrelated to body processing. Like human faces, human bodies activate specific cortical areas - namely the EBA and FBA - which may share features with the face-specific OFA and FFA. In addition, both faces and bodies activate the STS. Haxby et al. (2000) have suggested that face perception activates a core system (OFA, FFA and STS) which processes invariant face aspects and an extended system which processes variable facial features (see 4.1). Perception of human body shapes activates a neuronal network including the EBA, FBA and STS (Taylor et al., 2007). This network analyzes the visual appearance of bodies and is similar to the face-specific core system. This assumption is in agreement with the model proposed by Urgesi et al. (2004; 2007a). If further information about intentions or emotions of other people needs to be extracted from human body posture, additional brain regions are recruited which suggests further parallels to the extended system described by Haxby et al. (2000). For example, the mirror-neuron system is important for action perception and execution (see 3.1) as well as perception of others and personal emotions (see 3.2). A similar assumption is reported by the model described by Urgesi et al. (2007), which proposes two

STUDIEN

UND


dissociable networks for the perception of body forms and the perception of body movements or actions. A recent study of a double dissociation between the performance of patients with focal lesions in the EBA and ventral premotor cortex provides important evidence for the dissociation of these networks (Moro et al., 2008). In contrast to the frequently described selective face perception deficits (prosopagnosia), a comparable deficit in body perception has so far not been reported. However, imaging studies indicated close proximity between the face-specific FFA and the body-specific FBA and some clinical studies reported body-processing deficits in prosopagnosia subjects. As mentioned above (3.3), these studies did not include a broad range of neuropsychological data which are needed for the evaluation as to whether prosopagnosics have comparable face and body processing deficits. Alternatively, selective body processing deficits might be compensated by intact body processing areas or masked by greater difficulties with face perception (Peelen and Downing, 2007a). Moro et al. (2008) have recently described body form agnosia associated with EBA lesions, but the precise nature of the underlying impairment remains to be determined. The sparse data base on human body shape perception does not yet allow any firm conclusions as to the neuropsychological mechanisms contributing to the skill. Further studies are needed to clarify potential dissociations between face and body perception deficits and the degree of overlap in the underlying processing mechanisms.

STUDIEN

UND


References Allison, T., Puce, A., McCarthy, G., 2000. Social perception from visual cues: role of the STS region. Trends in Cognitive Science 4, 267. American Psychiatric Association, 1994. Diagnostic and statistical manual of mental disorders (4th edition). Washington, DC. Astafiev, S.V., Stanley, C.M., Shulman, G.L., Corbetta, M., 2004. Extrastriate body area in human occipital cortex responds to the performance of motor actions. Nature Neuroscience 7, 542. Behrmann, M., Avidan, G., 2005. Congenital prosopagnosia: face-blind from birth. Trends in Cognitive Science 9,180. Behrmann, M., Avidan, G., Marotta, J.J., Kimchi, R., 2005. Detailed exploration of face-related processing in congenital prosopagnosia: 1. Behavioral findings. Journal of Cognitive Neuroscience 17, 1130. Bentin, S., Allison, T., Puce, A., Perez, E., McCarthy, G., 1996. Electrophysiological studies of face perception in humans. Journal of Cognitive Neuroscience 8, 551. Bentin, S., Deouell, L.Y., Soroker, N., 1999. Selective visual streaming in face recognition: evidence from developmental prosopagnosia. Neuroreport 10, 823. Blake, R., Shiffrar, M., 2007. Perception of human motion. Annual Review of Psychology 58, 47. Blanke, O., Landis, T., Spinelli, L., Seeck, M., 2004. Out-of-body experience and autoscopy of neurological origin. Brain 127, 243. Blanke, O., Mohr, C., Michel, C.M., Pascual-Leone, A., Brugger, P., Seeck, M., Landis, T., Thut,G., 2005. Linking out-of-body experience and self processing to mental own-body imagery at the temporoparietal junction. Journal of Neuroscience 25, 550. Bodamer, J., 1947. Die Prosopagnosie. Archiv für Psychiatrie und Nervenkrankheiten 179, 6. Bosbach, S., Knoblich, G., Reed, C.L., Cole, J., Prinz, W., 2006. Body inversion effect without body sense: insights from deafferentation. Neuropsychologia 44, 2950.

STUDIEN

UND


Campanella, S., Belin, P., 2007. Integrating face and voice in person perception. Trends in Cognitive Science 11, 535. Chan, A.W., Peelen, M.V., Downing, P.E., 2004. The effect of viewpoint on body representation in the extrastriate body area. Neuroreport 15, 2407. Chao, L.L., Martin, A., Haxby, J.V., 1999. Are face-responsive regions selective only for faces? Neuroreport 10, 2945. Chong, T.T., Cunnington, R., Williams, M.A., Kanwisher, N., Mattingley, J.B., 2008. fMRI adaptation reveals mirror neurons in human inferior parietal cortex. Current Biology 18, 1576. Collishaw, S.M., Hole, G.J., 2000. Featural and configurational processes in the recognition of faces of different familiarity. Perception 29, 893. Cox, D., Meyers, E., Sinha, P., 2004. Contextually evoked object-specific responses in human visual cortex. Science 304, 115. Dailey, M.N., Cottrell, G.W., Padgett, C., Adolphs, R., 2002. EMPATH: a neural network that categorizes facial expressions. Journal of Cognitive Neuroscience 14, 1158. David, N., Cohen, M.X., Newen, A., Bewernick, B.H., Shah, N.J., Fink, G.R., Vogeleye, K., 2007. The extrastriate cortex distinguishes between the consequences of one‘s own and others‘ behavior. Neuroimage 36: 1004. Dawson, G., Webb, S.J., McPartland, J., 2005. Understanding the nature of face processing impairment in autism: insights from behavioral and electrophysiological studies. Developmental Neuropsychology 27, 403. de Gelder, B., 2006. Towards the neurobiology of emotional body language. Nature Reviews. Neuroscience 7, 242. de Gelder, B., Van den Stock, J., Balaguer, R.D., Bachoud-Levi, A.C., 2008. Huntington‘s disease impairs recognition of angry and instrumental body language. Neuropsychologia 46, 369. Diamond, R., Carey, S., 1986. Why faces are and are not special: an effect of expertise. Journal of Experimental Psychology. General 115, 107.

STUDIEN

UND


Dinstein, I., Thomas, C., Behrmann, M., Heeger, D.J., 2008. A mirror up to nature. Current Biology 18, R13. Downing, P.E., Chan, A.W., Peelen, M.V., Dodds, C.M., Kanwisher, N., 2006a. Domain specificity in visual cortex. Cerebral Cortex 16, 1453. Downing, P.E., Jiang, Y., Shuman, M., Kanwisher, N., 2001. A cortical area selective for visual processing of the human body. Science 293, 2470. Downing, P.E., Peelen, M.V., Wiggett, A.J., Tew, B.D., 2006b. The role of the extrastriate body area in action perception. Social Neuroscience 1, 52. Downing, P.E., Wiggett, A.J., Peelen, M.V., 2007. Functional magnetic resonance imaging investigation of overlapping lateral occipitotemporal activations using multi-voxel pattern analysis. Journal of Neuroscience 27, 226. Duchaine, B.C., Nakayama, K., 2006. Developmental prosopagnosia: a window to content-specific face processing. Current Opinion in Neurobiology 16, 166. Eimer, M., 2000a. Effects of face inversion on the structural encoding and recognition of faces. Evidence from event-related brain potentials. Brain Research. Cognitive Brain Research 10, 145. Eimer, M., 2000b. Event-related brain potentials distinguish processing stages involved in face perception and recognition. Clinical Neurophysiology 111, 694. Eimer, M., 2000c. The face-specific N170 component reflects late stages in the structural encoding of faces. Neuroreport 11, 2319. Eimer, M., Holmes, A., 2002. An ERP study on the time course of emotional face processing. Neuroreport 13, 427. Fairhall, S.L., Ishai, A., 2007. Effective connectivity within the distributed cortical network for face perception. Cerebral Cortex 17, 2400. Fox, C.J., Iaria, G., Barton, J.J., 2008. Disconnection in prosopagnosia and face processing. Cortex 44, 996. Gallese, V., Keysers, C., Rizzolatti, G., 2004. A unifying view of the basis of social cognition. Trends in Cognitive Science 8, 396.

STUDIEN

UND


Gauthier, I., Bukach, C., 2007. Should we reject the expertise hypothesis? Cognition 103, 322. Gauthier, I., Tarr, M.J., Anderson, A.W., Skudlarski, P., Gore, J.C., 1999. Activation of the middle fusiform ‚face area‘ increases with expertise in recognizing novel objects. Nature Neuroscience 2, 568. Gobbini, M.I., Haxby, J.V., 2007. Neural systems for recognition of familiar faces. Neuropsychologia 45, 32. Grezes, J., Pichon, S., de Gelder, B., 2007. Perceiving fear in dynamic body expressions. Neuroimage 35, 959. Grill-Spector, K., Knouf, N., Kanwisher, N., 2004. The fusiform face area subserves face perception, not generic within-category identification. Nature Neuroscience 7, 555. Hadjikhani, N., de Gelder, B., 2003. Seeing fearful body expressions activates the fusiform cortex and amygdala. Current Biology 13, 2201. Haxby, J.V., Hoffman, E.A., Gobbini, M.I., 2000. The distributed human neural system for face perception. Trends in Cognitive Science 4, 223. Haxby, J.V., Hoffman, E.A., Gobbini, M.I., 2002. Human neural systems for face recognition and social communication. Biological Psychiatry 51, 59. Haxby, J.V., Ungerleider, L.G., Clark, V.P., Schouten, J.L., Hoffman, E.A., Martin, A., 1999. The effect of face inversion on activity in human neural systems for face and object perception. Neuron 22, 189. Herrmann, M.J., Ehlis, A.C., Muehlberger, A., Fallgatter, A.J., 2005. Source localization of early stages of face processing. Brain Topography 18, 77. Iidaka, T., Matsumoto, A., Haneda, K., Okada, T., Sadato, N., 2006. Hemodynamic and electrophysiological relationship involved in human face processing: evidence from a combined fMRI-ERP study. Brain and Cognition 60, 176. Ishai, A., 2008. Let‘s face it: it‘s a cortical network. Neuroimage 40, 415. Ishai, A., Schmidt, C.F., Boesiger, P., 2005. Face perception is mediated by a distributed cortical network. Brain Research Bulletin 67, 87.

STUDIEN

UND


Itier, R.J., Taylor, M.J., 2004a. Effects of repetition learning on upright, inverted and contrast-reversed face processing using ERPs. Neuroimage 21, 1518. Itier, R.J., Taylor, M.J., 2004b. Face recognition memory and configural processing: a developmental ERP study using upright, inverted, and contrast-reversed faces. Journal of Cognitive Neuroscience 16, 487. Itier, R.J., Taylor, M.J., 2004c. N170 or N1? Spatiotemporal differences between object and face processing using ERPs. Cerebral Cortex 14, 132. Jacques, C., Rossion, B., 2007. Early electrophysiological responses to multiple face orientations correlate with individual discrimination performance in humans. Neuroimage 36, 863. Jokisch, D., Daum, I., Suchan, B., Troje, N.F., 2005. Structural encoding and recognition of biological motion: evidence from event-related potentials and source analysis. Behavioural Brain Research 157, 195. Kanwisher, N., McDermott, J., Chun, M.M., 1997. The fusiform face area: a module in human extrastriate cortex specialized for face perception. Journal of Neuroscience 17, 4302. Kanwisher, N., Yovel, G., 2006. The fusiform face area: a cortical region specialized for the perception of faces. Philosophical Transactions of the Royal Society of London. Series B, Biological Science 361, 2109. Kitada, R., Johnsrude, I.S., Kochiyama, T., Lederman, S.J., 2008. Functional specialization and convergence in the occipito-temporal cortex supporting haptic and visual identification of human faces and body parts: An fMRI Study. Journal of Cognitive Neuroscience (in press). Kress, T., Daum, I., 2003. Developmental prosopagnosia: a review. Behavioural Neurology 14, 109. Lange, J., Georg, K., Lappe, M., 2006. Visual perception of biological motion by form: a template-matching analysis. Journal of Vision 6, 836. Lange, J., Lappe, M., 2006. A model of biological motion perception from configural form cues. Journal of Neuroscience 26, 2894. Leder, H., Bruce, V., 2000. When inverted faces are recognized: the role of configural information in face recognition. The Quarterly Journal of Experimental Psychology A, Human Experimental Psychology 53, 513.

STUDIEN

UND


Lenggenhager, B., Tadi, T., Metzinger, T., Blanke, O., 2007. Video ergo sum: manipulating bodily selfconsciousness. Science 317, 1096. Marzi, T., Viggiano, M.P., 2007. Interplay between familiarity and orientation in face processing: An ERP study. International Journal of Psychophysiology 65, 182. Maurer, D., Grand, R.L., Mondloch, C.J., 2002. The many faces of configural processing. Trends in Cognitive Science 6, 255. McKone, E., Kanwisher, N., Duchaine, B.C., 2007. Can generic expertise explain special processing for faces? Trends in Cognitive Science 11, 8. Meeren, H.K., van Heijnsbergen, C.C., de Gelder, B., 2005. Rapid perceptual integration of facial expression and emotional body language. Proceedings of the National Academy of Sciences of the United States of America 102, 16518. Minnebusch, D.A., Suchan, B., Daum, I., 2008. Losing your head: behavioral and electrophysiological effects of body inversion. Journal of Cognitive Neuroscience (in press). Minnebusch, D.A., Suchan, B., Ramon, M., Daum, I., 2007. Event-related potentials reflect heterogeneity of developmental prosopagnosia. European Journal of Neuroscience 25, 2234. Moro, V., Urgesi, C., Pernigo, S., Lanteri, P., Pazzaglia, M., Aglioti, S.M., 2008. The neural basis of body form and body action agnosia. Neuron 60, 235. Morris, J.P., Pelphrey, K.A., McCarthy, G., 2006. Occipitotemporal activation evoked by the perception of human bodies is modulated by the presence or absence of the face. Neuropsychologia 44, 1919. Morris, J.S., Frith, C.D., Perrett, D.I., Rowland, D., Young, A.W., Calder, A.J., Dolan, R.J.,1996. A differential neural response in the human amygdala to fearful and happy facial expressions. Nature 383, 812. Norman, K.A., Polyn, S.M., Detre, G.J., Haxby, J.V., 2006. Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends in Cognitive Science 10, 424. Pavlova, M., Sokolov, A., 2000. Orientation specificity in biological motion perception. Perception & Psychophysics 62, 889.

STUDIEN

UND


Peelen, M.V., Downing, P.E., 2005. Selectivity for the human body in the fusiform gyrus. Journal of Neurophysiology 93, 603. Peelen, M.V., Downing, P.E., 2007a. The neural basis of visual body perception. Nature Reviews. Neuroscience 8, 636. Peelen, M.V., Downing, P.E., 2007b. Using multi-voxel pattern analysis of fMRI data to interpret overlapping functional activations. Trends in Cognitive Science 11, 4. Peelen, M.V., Wiggett, A.J., Downing, P.E., 2006. Patterns of fMRI activity dissociate overlapping functional brain areas that respond to biological motion. Neuron 49, 815. Pelphrey, K.A., Mitchell, T.V., McKeown, M.J., Goldstein, J., Allison, T., McCarthy, G., 2003. Brain activity evoked by the perception of human walking: controlling for meaningful coherent motion. Journal of Neuroscience 23, 6819. Pizzagalli, D.A., Lehmann, D., Hendrick, A.M., Regard, M., Pascual-Marqui, R.D., Davidson, R.J., 2002. Affective judgments of faces modulate early activity (approximately 160 ms) within the fusiform gyri. Neuroimage 16, 663. Pourtois, G., Schwartz, S., Seghier, M.L., Lazeyras, F., Vuilleumier, P., 2005. Portraits or people? Distinct representations of face identity in the human visual cortex. Journal of Cognitive Neuroscience 17, 1043. Puce, A., Allison, T., Bentin, S., Gore, J.C., McCarthy, G., 1998. Temporal cortex activation in humans viewing eye and mouth movements. Journal of Neuroscience 18, 2188. Puce, A., Perrett, D., 2003. Electrophysiology and brain imaging of biological motion. Philosophical Transactions of the Royal Society of London. Series B, Biological Science 358, 435. Reed, C.L., Stone, V.E., Bozova, S., Tanaka, J., 2003. The body-inversion effect. Psychological Science 14, 302. Reed, C.L., Stone, V.E., Grubb, J.D., McGoldrick, J.E., 2006. Turning configural processing upside down: part and whole body postures. Journal of Experimental Psychology. Human Perception and Performance 32, 73.

STUDIEN

UND


Reed, C.L., Beall, P.M., Stone, V.E., Kopelioff, L., Pulham, D.J., Hepburn, S.L., 2007. Brief report: perception of body posture-what individuals with autism spectrum disorder might be missing. Journal of Autism and Developmental Disorders 37, 1576. Righart, R., de Gelder, B., 2007. Impaired face and body perception in developmental prosopagnosia. Proceedings of the National Academy of Sciences of the United States of America 104, 17234. Rizzolatti, G., Craighero, L., 2004. The mirror-neuron system. Annual Review of Neuroscience 27, 169. Rizzolatti, G., Fogassi, L., Gallese, V., 2001. Neurophysiological mechanisms underlying the understanding and imitation of action. Nature Reviews. Neuroscience 2, 661. Rossion, B., 2008. Constraining the cortical face network by neuroimaging studies of acquired prosopagnosia. Neuroimage 40, 423. Rossion, B., Delvenne, J.F., Debatisse, D., Goffaux, V., Bruyer, R., Crommelinck, M., Guérit, J.M., 1999. Spatio-temporal localization of the face inversion effect: an event-related potentials study. Biological Psychology 50, 173. Rossion, B., Dricot, L., Devolder, A., Bodart, J.M., Crommelinck, M., de Gelder, B., Zoontjes, R., 2000a. Hemispheric asymmetries for whole-based and part-based face processing in the human fusiform gyrus. Journal of Cognitive Neuroscience 12, 793. Rossion, B., Gauthier, I., 2002. How does the brain process upright and inverted faces? Behavioral and Cognitive Neuroscience Reviews 1, 63. Rossion, B., Gauthier, I., Tarr, M.J., Despland, P., Bruyer, R., Linotte, S., Crommelinck, M., 2000b. The N170 occipito-temporal component is delayed and enhanced to inverted faces but not to inverted objects: an electrophysiological account of face-specific processes in the human brain. Neuroreport 11, 69. Schindler, K., Van Gool, L., de Gelder, B., 2008. Recognizing emotions expressed by body pose: A biologically inspired neural model. Neural Networks 21, 1238. Schwarzlose, R.F., Baker, C.I., Kanwisher, N., 2005. Separate face and body selectivity on the fusiform gyrus. Journal of Neuroscience 25, 11055.

STUDIEN

UND


Schweinberger, S.R., Pickering, E.C., Jentzsch, I., Burton, A.M., Kaufmann, J.M., 2002. Event-related brain potential evidence for a response of inferior temporal cortex to familiar face repetitions. Brain Research and Cognitive Brain Research 14, 398. Sergerie, K., Chochol, C., Armony, J.L., 2008. The role of the amygdala in emotional processing: a quantitative meta-analysis of functional neuroimaging studies. Neuroscience and Biobehavioral Reviews 32, 811. Stekelenburg, J.J., de Gelder, B., 2004. The neural correlates of perceiving human bodies: an ERP study on the body-inversion effect. Neuroreport 15, 777. Sumi, S., 1984. Upside-down presentation of the Johansson moving light-spot pattern. Perception 13, 283. Tanaka, J.W., Curran, T., Porterfield, A.L., Collins, D., 2006. Activation of preexisting and acquired face representations: the N250 event-related potential as an index of face familiarity. Journal of Cognitive Neuroscience 18, 1488. Tanaka, J.W., Farah, M.J., 1993. Parts and wholes in face recognition. The Quarterly Journal of Experimental Psychology A, Human Experimental Psychology 46, 225. Taylor, J.C., Wiggett, A.J., Downing, P.E., 2007. fMRI analysis of body and body part representations in the extrastriate and fusiform body areas. Journal of Neurophysiology 98, 1626. Teunisse, J.P., de Gelder, B., 2003. Face processing in adolescents with autistic disorder: the inversion and composite effects. Brain and Cognition 52, 285. Thierry, G., Pegna, A.J., Dodds, C., Roberts, M., Basan ,S., Downing, P., 2006. An event-related potential component sensitive to images of the human body. Neuroimage 32, 871. Thompson, J.C., Clarke, M., Stewart, T., Puce, A., 2005. Configural processing of biological motion in human superior temporal sulcus. Journal of Neuroscience 25, 9059. Thompson, J.C., Hardee, J.E., Panayiotou, A., Crewther, D., Puce, A. (2007). Common and distinct brain activation to viewing dynamic sequences of face and hand movements. Neuroimage 37, 966. Troje, N.F., Westhoff, C., 2006. The inversion effect in biological motion perception: evidence for a “life detector”? Current Biology 16, 821.

STUDIEN

UND


Urgesi, C., Berlucchi, G., Aglioti, S.M., 2004. Magnetic stimulation of extrastriate body area impairs visual processing of nonfacial body parts. Current Biology 14, 2130. Urgesi, C., Calvo-Merino, B., Haggard, P., Aglioti, S.M., 2007a. Transcranial magnetic stimulation reveals two cortical pathways for visual body processing. Journal of Neuroscience 27, 8023. Urgesi, C., Candidi, M., Ionta, S., Aglioti, S.M., 2007b. Representation of body identity and body actions in extrastriate body area and ventral premotor cortex. Nature Neuroscience 10, 30. Van den Stock, J., van de Riet, W.A., Righart, R., de Gelder, B., 2008. Neural correlates of perceiving emotional faces and bodies in developmental prosopagnosia: an event-related fMRI-study. PLoS ONE 17, e3195. van Heijnsbergen, C.C., Meeren, H.K., Grezes, J., de Gelder, B., 2007. Rapid detection of fear in body expressions, an ERP study. Brain Research 1186, 233. Vuilleumier, P., Pourtois, G., 2007. Distributed and interactive brain mechanisms during emotion face perception: evidence from functional neuroimaging. Neuropsychologia 45, 174. Yin, R.K., 1968. Looking at upside- down faces. Journal of Experimental Psychology 81, 141. Yin, R.K., 1970. Face recognition by brain-injured patients: a dissociable ability? Neuropsychologia 8, 395. Young, A.W,, Hellawell, D., Hay, D.C., 1987. Configurational information in face perception. Perception 16, 747.

GESAMTDISKUSSION und AUSBLICK I DREI SEITE 115

3. GESAMTDISKUSSION und AUSBLICK Die Mechanismen der Verarbeitung menschlicher Gesichter stehen seit langer Zeit im Fokus neurokognitiver Forschungsinteressen. Hinweise auf ein selektives Gesichterverarbeitungsdefizit (Prosopagnosie) bei gleichzeitig intakter Objektverarbeitung belegen die Annahme, dass Gesichter teilweise anders verarbeitet werden als andere Objekte. Obgleich es Anzeichen für die Existenz eines neuronalen Gesichterverarbeitungsnetzwerkes - bestehend aus der FFA und OFA - gibt, ist bislang unklar, wie die einzelnen Areale dieses Netzwerks miteinander interagieren und ob diesem Netzwerk weitere Areale zuzuordnen sind. Der Inversionseffekt liefert einen weiteren Beleg für die Sonderstellung menschlicher Gesichter innerhalb der visuellen Reizverarbeitung (Yin, 1969; 1970) und wird als Indikator für konfigurale Verarbeitungsprozesse angesehen (Leder et al., 2001; Leder & Bruce, 2000; Maurer et al., 2002). Ein vergleichbarer Inversionseffekt konnte auch für die Verarbeitung menschlicher Körper nachgewiesen werden, was auf die Beteilung ähnlicher Prozesse bei der Gesichter- und Körperverarbeitung hindeutet (Reed et al., 2003; 2006). Die vorliegende Arbeit setzt sich aus vier empirischen Studien sowie einem Übersichtsartikel zusammen und beschäftigt sich mit Gemeinsamkeiten sowie Unterschieden der neurokognitiven Verarbeitungsmechanismen menschlicher Gesichter und Körper. Der Fokus der ersten empirischen Studie lag auf der Untersuchung möglicher Ursachen der angeborenen Prosopagnosie, wobei sich die anderen drei Studien mit den Mechanismen der Körperverarbeitung beschäftigten. Hierbei stand vor allem die Frage einer möglichen konfiguralen Körperverarbeitung im Zentrum des Interesses. Die Frage nach der Entstehung des Körperinversionseffektes stellte ein weiteres zentrales Ziel der Arbeit dar. Zudem sollte untersucht werden, bis zu welchem Verarbeitungsschritt Parallelen in der Gesichter- und Körperwahrnehmung nachgewiesen werden können. Eine Zusammenfassung und Diskussion der zentralen Befunde der Arbeit erfolgt in den nachfolgenden Abschnitten, wobei gesondert auf die zuvor definierten Ziele eingegangen wird.


3.1

Neuronale Mechanismen der Gesichterverarbeitung

Die Verarbeitung menschlicher Gesichter nimmt innerhalb der visuellen Wahrnehmung eine Sonderstellung ein. Die Existenz eines selektiven Gesichterverarbeitungsdefizits (Prosopagnosie) bei intakter Objektverarbeitung gilt als einer der stärksten Belege für diese Annahme. Die erworbene Prosopagnosie kann von der angeborenen Variante abgegrenzt werden. Die angeborene Prosopagnosie ist sehr selten und bislang ist unklar, welche neuronalen Mechanismen diesem Defizit zugrunde liegen. Ziel der ersten Studie war es, die neuronalen Mechanismen der intakten Gesichterverarbeitung mittels fMRT und PPI-Analyse genauer zu untersuchen (Minnebusch et al., 2008b). Auf der Grundlage dieser Befunde sollten die Defizite der Prosopagnostiker besser verstanden werden. Zudem sollten die gesichtsspezifischen Aktivierungen gesunder Kontrollprobanden mit denen von angeborenen Prosopagnistikern verglichen werden. Die Ergebnisse dieser Untersuchung unterstützen die Annahme, dass Gesichter ein weit verbreitetes kortikales Netzwerk aktivieren (Haxby et al., 2000). Mit Hilfe einer PPI-Analyse haben wir Regionen identifiziert, die während der Gesichterverarbeitung mit der rechten FFA interagieren. Neben der linken FFA und der rechten und linken OFA scheinen weitere Areale von Bedeutung zu sein. In diesem Zusammenhang sind vor allem visuelle Regionen wie der Cuneus, der linguale Gyrus und der mittlere okzipitale Gyrus zu nennen. Welche Funktion diesen Arealen bei der Gesichterverarbeitung zukommt, muss in weiteren Studien geklärt werden (z.B. DCM, „Diffusion Tensor Imaging“). Vorherige Untersuchungen (Fairhall & Ishai, 2007; Haxby et al., 2000; Ishai, 2008; Rossion et al., 2003; Rossion, 2008) sowie Befunde der ersten Studie (Minnebusch et al., 2009a) stützen die Annahme, dass die Integrität der gesichtssensitiven Regionen (rechte und linke FFA sowie OFA) für die erfolgreiche Gesichterverarbeitung notwendig ist. Die gefundenen heterogenen Aktivierungsmuster innerhalb der Gruppe der Prosopagnostiker stehen in Übereinstimmung mit zuvor beobachteten Verhaltensmustern (Le Grand et al., 2006) und gesichtsevozierten elektrophysiologischen Potentialen (Harris et al., 2005; Minnebusch et al., 2007). Hervorzuheben ist, dass Gesichter bei keinem der Probanden mit Prosopagnosie bilaterale Aktivierungen in der FFA und OFA ausgelöst haben. Zudem zeigte sich ein Zusammenhang zwischen den Beeinträchtigungen der Patienten und den aktivierten Hirnarealen. Die Ergebnisse können im Sinne eines gesichtsspezifischen kortikalen Netzwerkes interpretiert werden, dessen gemeinsame Aktivität für die erfolgreiche Gesichterverarbeitung notwendig ist. Im Fokus dieses Netzwerkes stehen die FFA und OFA, wobei normale Aktivierungen in diesen Bereichen nicht ausreichend für eine erfolgreiche Gesichterverarbeitung sind. Die Heterogenität der Prosopagnosie kann darauf zurückgeführt werden, dass selektive Gesichterverarbeitungsdefizite auftreten können, wenn in nur einem Areal dieses Netzwerkes eine Fehlfunktion vorliegt. Wiederum scheint die funktionelle Integrität der FFA und OFA in diesem Zusammenhang von besonderer Bedeutung zu sein.


3.2

Konfigurale Körperverarbeitung

Die Befunde der zweiten, dritten und vierten empirischen Studie deuten darauf hin, dass konfigurale Verarbeitungsprozesse auch bei der Körperverarbeitung von Bedeutung sind. Zudem scheint die N170 ein neuronales Korrelat des Inversionseffektes darzustellen, sodass angenommen werden kann, dass die konfigurale Verarbeitung 170 ms nach der Reizwahrnehmung auftritt. Vorherige Studien sowie die Ergebnisse der vorliegenden Arbeit zeigen, dass menschliche Körper, ähnlich wie menschliche Gesichter, eine Sonderstellung innerhalb der visuellen Reizverarbeitung einnehmen. Erste Hinweise darauf, dass Körper zumindest zum Teil anders verarbeitet werden als andere Objekte stammen von Studien zum Inversionseffekt, welcher als ein Indikator für konfigurale Verarbeitungsprozesse angesehen wird (Reed et al., 2003; 2006; Stekelenburg & de Gelder, 2004). Der Inversionseffekt wurde zuvor nur bei der Gesichterverarbeitung beobachtet und stellt ein sehr robustes und reliables Phänomen dar (Farah et al., 1995; Freire et al., 2000; Jacques & Rossion, 2007; Yin, 1969; 1970). Vertreter der konfiguralen Verarbeitungshypothese vermuten die Ursache des Inversionseffektes vorrangig in der Verarbeitung der räumlichen Relationen zwischen einzelnen Gesichts- bzw. Körpermerkmalen (Leder et al., 2001; Leder & Bruce, 2000; Maurer et al., 2002). Des Weiteren unterscheiden sich die EKP aufrechter Gesichter und Körper von jenen, welche bei der Wahrnehmung invertierter Gesichter/Körper ausgelöst werden (Bentin et al.,1996; Eimer, 2000a; Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004). Dieser Effekt äußert sich hauptsächlich in einer verzögerten und zum Teil erhöhten Amplitude der N170 bei invertierter verglichen mit aufrechter Reizverarbeitung. In der Gesichterliteratur herrscht Einigkeit darüber, dass die N170 mit der strukturellen Reizenkodierung assoziiert ist und ein neuronales Korrelat der konfiguralen Verarbeitung darstellt. In den beiden EEG-Studien, welche den Inversionseffekt bei menschlichen Körpern untersucht haben, zeigten die Autoren Körper mit maskierten Gesichtern als Stimuli (Righart & de Gelder, 2007; Stekelenburg & de Gelder, 2004). Unklar ist bislang, ob die beobachteten Effekte durch die Körper, die maskierten Gesichter oder beides ausgelöst wurden. In der zweite Studie (Minnebusch et al., 2008) wurde der Inversionseffekt bei Körpern mit und ohne Kopf sowie Tieren mittels Verhaltensmaßen und EKPs untersucht. Ziel der Studie war zu untersuchen, ob die beobachteten Inversionseffekte auf die maskierten Gesichter zurückgeführt werden können. Ein weiteres Ziel dieser Studie bestand darin zu überprüfen, ob die Inversionseffekte nur bei menschlichen Körpern auftreten oder ob eine Generalisierung auf alle Körper stattfindet. In dieser Studie konnte der Inversionseffekt für Körper mit maskierten Gesichtern sowohl auf der Verhaltensebene als auch auf elektrophysiologischer Ebene repliziert werden. Ein umgekehrtes Befundmuster (bessere Leistungen und kleinere N170-Amplituden bei invertierten verglichen mit aufrechten Körpern) zeigte sich für die Kategorie Körper ohne Kopf. Diese Befunde deuten darauf hin, dass Körper konfigural verarbeitet werden, wobei Körper ohne Kopf (ähnlich wie andere Objekte) auf der Basis einzelner Merkmale verarbeitet werden.


Ein alternativer Erklärungsansatz bezieht sich auf die unnatürliche Darstellung der Körper ohne Kopf, welche möglicherweise nicht genügend Hinweisreize liefern, um das kortikale Körperschema zu aktivieren. Diese Reize werden demnach nicht konfigural verarbeitet, weil sie nicht als Körper erkannt werden. Des Weiteren besteht die Möglichkeit, dass die berichteten Effekte für die Körper mit Kopf auf die Darstellung des Kopfes zurückzuführen sind, der gesichtsspezifische Verarbeitungsmechanismen aktiviert. Zusammenfassend ist festzuhalten, dass Körper mit Kopf vermutlich anders verarbeitet werden als Körper ohne Kopf. Die genaue Ursache dieses Effektes sollte in weiteren Studien mittels verschiedener Paradigmen untersucht werden. Die zuvor berichteten Unterschiede zwischen Körpern mit und ohne Kopf konnten in den nachfolgenden Studien nur teilweise repliziert werden (Studie 3 und 4). In diesem Zusammenhang sollte berücksichtigt werden, dass in diesen Experimenten andere Stimuli und in der vierten Studie zudem ein anderer experimenteller Kontext verwendet wurden. Trugen die gezeigten Personen der zweiten Studie (Minnebusch et al., 2008) noch unterschiedliche Kleidung, zeigten wir in den folgenden Studien Körperstimuli, welche annähernd die gleiche Kleidung trugen (Minnebusch, et al., 2009b; Minnebusch & Daum, 2009a). In der dritten Studie konnte für beide Reizkategorien ein quadratischer Zusammenhang zwischen dem Rotationswinkel und dem Verhaltensmuster sowie dem Rotationswinkel und den elektrophysiologischen Korrelaten (P100, N170) beobachtet werden. Die konfigurale Körperverarbeitung findet wie die konfigurale Gesichterverarbeitung während der strukturellen Reizenkodierung statt. In der vierten Studie (Minnebusch & Daum, 2009a) fanden wir für beide Reizkategorien einen behavioralen Inversionseffekt, jedoch wiederum stärkere Unterschiede für Körper mit Kopf verglichen mit Körpern ohne Kopf. An dieser Stelle sei jedoch darauf hinzuweisen, dass die Aufmerksamkeit der Probanden in dieser Studie aktiv nur auf die Oberkörper gelenkt wurde. Berücksichtigt man in diesem Zusammenhang die fehlenden Hinweise auf eine Beteiligung holistischer Verarbeitungsprozesse bei der Wahrnehmung von Körpern ohne Kopf kann nicht ausgeschlossen werden, dass in dieser Studie nur Inversionseffekte bezüglich der Verarbeitung des Oberkörpers anstatt der Effekte auf die Verarbeitung des gesamten Körpers gemessen wurden. Insgesamt lässt sich festhalten, dass Körper mit Köpfen, wie Gesichter, konfigural verarbeitet werden und dass dieser Prozess während der strukturellen Enkodierung des Reizes auftritt. Bezüglicher der Verarbeitung von Körpern ohne Kopf liegen widersprüchliche Befunde vor, so dass weitere Studien nötig sind, um die Bedeutung der vorliegenden Ergebnisse genauer abzuklären.


Die Frage nach der Generalisierung des Inversionseffektes auf alle Körper kann aufgrund des heterogenen Musters nicht eindeutig beantwortet werden (Minnebusch et al., 2008). Die vorliegenden Befunde deuten jedoch darauf hin, dass Tiere und menschliche Körper eine vergleichbare N170 evozieren, welche sich signifikant von jener Komponente unterscheidet, welche durch Objekte (hier Häuser) ausgelöst wird. Zudem konnte sowohl auf der Verhaltens- als auch auf der elektrophysiologischen Ebene ein Inversionseffekt nachgewiesen werden, wenngleich kein systematischer Zusammenhang zwischen beiden Mustern zu finden war: Hunde lösten auf der Verhaltensebene einen Inversionseffekt aus, nicht aber Katzen, für welche hingegen ein elektrophysiologischer Inversionseffekt beobachtet wurde. Einheitlich war das Muster hingegen für die Kategorie Vögel, für die auf beiden Ebenen keine Unterschiede zwischen der aufrechten und invertierten Verarbeitung gefunden wurden. Über mögliche Ursachen dieser Effekte kann bislang nur spekuliert werden. Zum einen besteht die Möglichkeit, dass Tiere, menschliche Körper und Gesichter auf unterschiedlichen Ebenen des konfiguralen Verarbeitungskontinuums (siehe Abbildung 1; Maurer et al., 2002; Reed et al., 2006) einzuordnen sind. Während sich Gesichter, für deren Verarbeitung alle Ebenen wichtig sind, am äußersten rechten Ende befinden, sind Körper in der Mitte (strukturelle Informationen oder holistische Verarbeitung) und Tiere am Anfang des Kontinuums lokalisiert (siehe Abbildung 9). In anderen Worten, „first-order relations“ könnten für die Wahrnehmung aller drei Kategorien (Tiere, Körper, Gesichter) von Bedeutung sein, wohingegen strukturelle Informationen und eventuell holistische Verarbeitungsmechanismen nur für die Wahrnehmung menschlicher Körper und Gesichter notwendig sind. „Second-order relations“, welche am rechten Ende des Kontinuums lokalisiert sind, spielen nur für die Gesichterverarbeitung eine Rolle. Alternativ kann angenommen werden, dass die beobachteten Befunde auf die klassische Unterteilung in belebte und unbelebte Objekte und auf damit zusammenhängende Effekte zurückzuführen sind. Die berichteten Aktivierungsunterschiede in frühen Potentialen (z.B. N170) unterstützen diese Annahme. Zusammenfassend ist festzuhalten, dass die Bedeutung der vorliegenden Befunde in nachfolgenden Studien genauer geklärt werden muss. Die Annahme, dass Tiere, deren Körperform der menschlichen Körperform ähnelt, konfigural verarbeitet werden, konnte nicht bestätigt werden.


Abbildung 9: Modifiziertes Modell der Objekt-, Körper- und Gesichterverarbeitung (Maurer et al., 2002). Die Verarbeitung einzelner Stimuluskomponenten ist für alle drei Kategorien wichtig. Strukturellen und eventuell auch holistischen Verarbeitungsmechanismen kommt bei der Gesichter- und Körperverarbeitung eine zentrale Funktion zu. Die Bedeutung von „second-order relations“ konnte bislang nur für die Gesichterwahrnehmung nachgewiesen werden.

3.3

Parallelen und Unterschiede der Gesichter- und Körperwahrnehmung

In dem folgenden Abschnitt werden die beobachteten Gemeinsamkeiten und Unterschiede bezüglich der Gesichter- und Körperverarbeitung zusammengefasst und kritisch diskutiert. Mit dieser Frage hat sich sowohl der Übersichtsartikel (Minnebusch & Daum, 2009b) als auch die Mehrzahl der vorliegenden empirischen Studien (Studie 2-4) beschäftigt. Elektrophysiologische Studien, Bildgebungsstudien und Verhaltensexperimente berichten Gemeinsamkeiten zwischen der visuellen Gesichter- und Körperverarbeitung. Das Fehlen eines selektiven Körperverarbeitungsdefizits, analog der Prosopagnosie, stellt zum jetzigen Zeitpunkt den größten Unterschied zwischen beiden Kategorien dar.

3.3.1

Parallelen der Gesichter- und Körperwahrnehmung

Gesichter und Körper liefern eine Vielzahl sozialer Hinweisreize, welchen in der zwischenmenschlichen Kommunikation und Interaktion eine große Bedeutung zukommt. In der Literatur herrscht Einigkeit darüber, dass Gesichter innerhalb der visuellen Wahrnehmung eine Sonderstellung einnehmen. Gesichter werden in selektiven kortikalen Arealen und mittels spezifischer Mechanismen verarbeitet, welche sich von Mechanismen der Objektverarbeitung unterscheiden. Objekt- und Gesichterverarbeitung basieren auf der Wahrnehmung einzelner Reizkomponenten. Letztere werden zudem konfigural verarbeitet, was im


experimentellen Kontext primär mittels Stimulusinversion untersucht wurde (Yin, 1969; 1970). Ein vergleichbarer Inversionseffekt wurde auch für menschliche Körper nachgewiesen, was ein erster Hinweis darauf ist, dass Körper innerhalb der visuellen Wahrnehmung auch eine Sonderstellung einnehmen (Minnebusch et al., 2008; Reed et al., 2003; 2006). Unklar bleibt in diesem Zusammenhang jedoch, ob konfigurale Verarbeitungsprozesse auf die Wahrnehmung des Körpers, des Kopfes oder des Körpers und des Kopfes zurückzuführen sind (siehe 3.2). Gesichter und Körper werden in benachbarten und zum Teil überlappenden kortikalen Arealen verarbeitet, deren anatomische Lokalisation sich wiederum von objektspezifischen Regionen unterscheidet (Kanwisher & Yovel, 2006; Peelen & Downing, 2007a; Schwarzlose et al., 2005). Im Gyrus fusiformis konnten distinkte gesichtsund körpersensitive Areale nur mittels hochauflösender Messverfahren identifiziert werden (Schwarzlose et al., 2005). Herkömmliche bildgebende Verfahren deuteten zunächst darauf hin, dass Gesichter und Körper in dieser Region in identischen Arealen verarbeitet werden. Weitere Parallelen bestehen darin, dass Gesichter und Körper in der OFA/EBA auf der Basis einzelner Details und im fusiformen Gyrus (FFA/FBA) konfigural verarbeitet werden (Fairhall & Ishai, 2007; Haxby et al., 2000; Ishai, 2008; Taylor et al., 2007). Gesichtsspezifische Informationen werden von der OFA in die FFA geleitet (Fairhall & Ishai, 2007). Des Weiteren existieren Feedbackprojektionen von der FFA in die OFA (Rossion, 2008) und Interaktionen zwischen der rechten FFA und der linken FFA sowie der rechten und linken OFA (Minnebusch et al., 2009a). Die Interaktionen zwischen den körpersensitiven Regionen wurden bislang nicht untersucht. Auf der Grundlage vorheriger Studien sowie der berichteten Parallelen in der Gesichter- und Körperverarbeitung ist anzunehmen, dass körperspezifische Informationen zunächst in der EBA auf der Basis einzelner Details verarbeitet werden. Von dort gelangen die Informationen in die FBA, innerhalb derer eine konfigurale Verarbeitung stattfindet. Allerdings fehlen bislang Hinweise auf konfigurale Verarbeitungsprozesse innerhalb der FBA (siehe Minnebusch & Daum, 2009b). Geringere FFA-Aktivierungen bei der Wahrnehmung invertierter verglichen mit aufrechten Gesichtern, welche mit entsprechenden Verhaltensdaten korrelieren, belegen die Bedeutung konfiguraler Verarbeitungsprozesse innerhalb dieser Region (Yovel & Kanwisher, 2005). Vergleichbare Untersuchungen könnten Hinweise darauf liefern, ob Körper in der FBA konfigural verarbeitet werden. Eine weitere Gemeinsamkeit besteht darin, dass beide Reizkategorien vergleichbare elektrophysiologische Potentiale (P100, N170) evozieren, welche sich signifikant von objektspezifischen Potentialen unterscheiden (Minnebusch et al., 2008a; Rossion et al., 2000b; Thierry et al., 2006). Zudem findet die konfigurale Gesichter- und Körperverarbeitung annährend 170 ms nach der Reizwahrnehmung statt (Jacques & Rossion, 2007; Minnebusch et al., 2008). Betrachtet man die einzelnen Ebenen des konfiguralen Verarbeitungskontinuums deuten die aktuellen Befunde darauf hin, dass „first-order relations“ und strukturelle Informationen für beide Reizklassen von Bedeutung sind (Maurer et al., 2002; Reed et al., 2003; 2006). Holistischen Verarbeitungsprozessen scheint bei der Wahrnehmung menschlicher


Gesichter eine zentralere Funktion zuzukommen als dies bei der Körperwahrnehmung der Fall ist, wie Studie 4 der vorliegenden Arbeit anhand des „composite effects“ zeigte (Minnebusch & Daum, 2009a). Studien, welche die Bedeutung von „second-order relations“ für die Körperverarbeitung untersuchen, fehlen bislang. Neuere Studien deuten zudem darauf hin, dass Personen mit Prosopagnosie neben Defiziten bei der konfiguralen Gesichterverarbeitung auch eine defizitäre konfigurale Körperverarbeitung aufweisen (Righart & de Gelder, 2007). Allerdings basiert diese Annahme ausschließlich auf elektrophysiologischen Daten. Van den Stock et al. (2008) berichten, dass bei Prosopagnostikern eine reduzierte kategoriale Differenzierung gesichtsund körpersensitiver Mechanismen vorliegt. Weitere Studien, welche Verhaltensdaten, elektrophysiologische und möglicherweise auch bildgebende Daten kombinieren, müssen die Bedeutung dieser Befunde genauer klären.

3.3.2

Unterschiede zwischen der Gesichter- und Körperwahrnehmung

Die Ergebnisse der Literaturanalyse ergaben, dass das Fehlen eines der Prosopagnosie vergleichbaren Körperverarbeitungsdefizits bislang den größten Unterschied zwischen beiden Reizklassen darstellt (Minnebusch & Daum, 2009b). Zwar postulieren Moro et al. (2008) die Existenz einer erworbenen Körperformagnosie, welche mit EBA-Läsionen assoziiert ist, ohne jedoch nachzuweisen, dass diese Patienten Schwierigkeiten mit der Verarbeitung von Körpern haben. Eine umfassende neuropsychologische Charakterisierung sowie Vergleiche zwischen diesen Patienten und Personen mit Prosopagnosie sind für den Nachweis eines derartigen Verarbeitungsdefizits notwendig. In dem vorherigen Abschnitt wurden Parallelen in den neuroanatomischen und elektrophysiologischen Korrelaten der Gesichter- und Körperverarbeitung dargestellt (siehe 3.4.1). Trotz dieser Gemeinsamkeiten handelt es sich nicht um identische neuronale Marker. Thierry und Mitarbeiter (2006) identifizierten unterschiedliche neuronale Quellen für die gesichtssensitive N170 und körpersensitive N170 bzw. N190. Des Weiteren ist die gesichtssensitive OFA deutlich von der körpersensitiven EBA abzugrenzen und auch innerhalb des fusiformen Gyrus konnten distinkte Areale für die Gesichter- und Körperverarbeitung identifiziert werden. Zusammenfassend ist festzuhalten, dass zahlreiche Parallelen aber auch Unterschiede zwischen der Gesichter- und Körperverarbeitung existieren. Die beteiligten Prozesse können von denen abgegrenzt werden, die für die Verarbeitung anderer Objekte zuständig sind. Die geringe Anzahl an Studien, welche sich mit den Mechanismen der Körperverarbeitung beschäftigt, lässt zum jetzigen Zeitpunkt keine definitiven Schlussfolgerungen zu.


3.4

Ausblick

Verglichen mit der Vielzahl an Studien, welche die Mechanismen der Gesichterverarbeitung untersucht haben, sind die derzeitigen Erkenntnisse zu Mechanismen der Körperverarbeitung als unzureichend einzustufen. Viele Fragen können aufgrund fehlender empirischer Daten bislang nicht beantwortet werden. Auf neuroanatomischer Ebene wurden zwei Regionen identifiziert, welche für die Körperverarbeitung von zentraler Bedeutung sind, die EBA und die FBA. Unklar ist bislang, in welcher Weise diese Regionen miteinander interagieren. Des Weiteren stellt sich die Frage, ob und wie diese Regionen mit dem gesichtsspezifischen Netzwerk, insbesondere der FFA und OFA, kommunizieren. Gesichter und Körper werden im nicht experimentellen Kontext (im Alltag) als Ganzes wahrgenommen und liefern soziale Hinweisreize, welche für die zwischenmenschliche Interaktion von großer Bedeutung sind. Aufgrund dessen sind die berichteten Gemeinsamkeiten in der kortikalen Verarbeitung beider Reizklassen nicht verwunderlich. Untersuchungen mittels DTI könnten neue Erkenntnisse über die Faserverbindungen zwischen gesichtsund körperspezifischen Regionen liefern. Des Weiteren sollten DCM- und PPI-Analysen die Frage klären, welche Interaktionen zwischen diesen Arealen bestehen und ob weitere Regionen für die erfolgreiche Körperverarbeitung notwendig sind. Auf der Grundlage des Körperverarbeitungsmodells nach Taylor et al. (2007) kann die Hypothese formuliert werden, dass Körper zunächst in der EBA auf der Basis einzelner Details und anschließend in der FBA konfigural verarbeitet werden. Vergleichbare EEG-Studien könnten zudem klären, zu welchem Zeitpunkt diese Prozesse im Gehirn ablaufen. Darüber hinaus wäre es interessant zu untersuchen, ob die gesichtsund körperspezifischen Informationen ab einem bestimmten Verarbeitungsschritt zu einem Gesamtbild integriert werden und zu welchem Zeitpunkt dieser Zusammenschluss stattfindet. Möglicherweise findet diese Integration in den Bereichen des fusiformen Gyrus statt, indem eine Überlappung zwischen der körpersensitiven FBA und der gesichtssensitiven FFA beobachtet wurde (Schwarzlose et al., 2005). Mittels funktioneller Bildgebung sollten zudem die neuroanatomischen Korrelate des Körperinversionseffektes identifiziert und wenn möglich von den entsprechenden gesichtsspezifischen Regionen abgegrenzt werden. Neben der P100 und N170 wurde mit der N250 ein weiteres gesichtsspezifisches EKP identifiziert (Itier & Taylor, 2004a; Schweinberger et al., 2002; Tanaka et al., 2006), welches mit der Bekanntheit eines Gesichts assoziiert ist. Es wäre von Interesse, ob bekannte Körper ein vergleichbares Potential evozieren. Funktionelle Bildgebungsstudien berichteten außerdem Effekte der Menge an dargebotenen Körperinformationen: Körperteile und Körper rufen vergleichbare EBA-Aktivierungen hervor, wobei FBA-Aktivierungen linear mit der Menge der dargebotenen Körperinformationen ansteigen (Taylor et al., 2007). Daraus ergibt sich die Frage, ob die beobachteten elektrophysiologischen Potentiale auch durch einzelne


Körperteile evoziert werden. Möglicherweise treten bestimmte frühe Potentiale (z.B. P100) bereits bei der Darbietung einzelner Körperteile auf, wobei spätere Potentiale (z.B. N170) erst durch den ganzen Körper evoziert werden können. Eine weitere wichtige Zielsetzung künftiger Forschungsvorhaben besteht darin zu überprüfen, ob Läsionen innerhalb der EBA und/oder FBA ein selektives Körperverarbeitungsdefizit (Körperformagnosie) verursachen. Im Mittelpunkt des Interesses sollte hierbei eine umfangreiche neuropsychologische Diagnostik stehen, mittels derer eine Objektagnosie sowie eine Prosopagnosie ausgeschlossen werden können. Die Entwicklung einer zuverlässigen Diagnosebatterie, mit welcher Körperverarbeitungsdefizite nachgewiesen werden können, ist eine notwendige Voraussetzung für die Durchführung derartiger Studien. Bislang ist zudem wenig über mögliche Körperverarbeitungsdefizite bei Personen mit Prosopagnosie bekannt. Das Vorliegen einer Kombination aus Prosopagnosie und Körperformagnosie könnte ein Beleg dafür sein, dass beide Reizkategorien mittels identischer Mechanismen verarbeitet werden. Die anatomische Nähe der gesichtsund körpersensitiven Areale, insbesondere im fusiformen Gyrus, muss jedoch bei der Interpretation der Ergebnisse berücksichtigt werden: Läsionen im fusiformen Gyrus könnten sowohl die FFA als auch die FBA betreffen und eine mögliche Ursache für das gemeinsame Auftreten derartiger Defizite darstellen. Auf der anderen Seite würde der Nachweis einer doppelten Dissoziation (d.h. Patienten mit einer Prosopagnosie und intakter Körperverarbeitung sowie Patienten mit einer Körperformagnosie und intakter Gesichterverarbeitung) die These stützen, dass Gesichter- und Körperverarbeitung auf unterschiedlichen Mechanismen beruhen. Die Befunde der vorliegenden empirischen Studien und die Ergebnisse der Literaturanalyse in dem Übersichtsartikel deuten darauf hin, dass Gesichter und Körper konfigural verarbeitet werden. Betrachtet man die einzelnen Ebenen des konfiguralen Verarbeitungskontinuums (siehe Abbildung 1) so ist anzunehmen, dass „first-order relations“ und strukturelle Informationen für beide Kategorien von Bedeutung sind. Die vierte Studie dieser Arbeit liefert erste Hinweise darauf, dass auch holistische Verarbeitungsprozesse an der Körperwahrnehmung beteiligt sind, wenngleich der Einfluss dieser Prozesse, verglichen mit der Gesichterverarbeitung, geringer zu sein scheint. Mittels weiterer Studien sollte die Bedeutung dieser Befunde weiter abgeklärt werden. Außerdem wäre es interessant zu untersuchen, ob „secondorder relations“ bei der Körperwahrnehmung eine wichtige Rolle spielen. Derartige Erkenntnisse würden weitere Hinweise dafür liefern, auf welcher Ebene des konfiguralen Verarbeitungskontinuums Körperstimuli einzuordnen sind (siehe Abbildung 1).

ZUSAMMENFASSUNG I VIER SEITE 125

4. ZUSAMMENFASSUNG Menschliche Gesichter und Körper übermitteln zahlreiche soziale Informationen, welche für die Identifikation anderer Personen, ihres Alters, Geschlechts sowie ihrer Absichten und Emotionen genutzt werden können. Die neuropsychologischen Mechanismen der Gesichterverarbeitung, die diesen Prozessen zugrunde liegen, sind seit langer Zeit Gegenstand neurowissenschaftlicher Forschungsinteressen. Neuere Studien beschäftigen sich mit den Prozessen, welche für die Körperwahrnehmung von zentraler Bedeutung sind. Erste Befunde deuten auf Gemeinsamkeiten in der Gesichter- und Körperverarbeitung hin, aber auch Unterschiede wurden berichtet. Ziel der vorliegenden Arbeit war es, neue Erkenntnisse über die neurokognitiven Mechanismen der Gesichterwahrnehmung zu gewinnen, die zu einem besseren Verständnis der möglichen Ursachen der Prosopagnosie beitragen sollten (Minnebusch et al., 2009a). Die Ergebnisse der ersten Studie deuten darauf hin, dass ein bilaterales Netzwerk bestehed aus der OFA und FFA die Grundlage für die erfolgreiche Gesichterverarbeitung ist. Dysfunktionen in einer dieser Regionen könnten zu Defiziten der Gesichterverarbeitung führen. Zudem wurden Parallelen sowie Unterschiede in der Verarbeitung menschlicher Gesichter und Körper untersucht (Studien 2-4). Außerdem wurden die bereits vorliegenden Erkenntnisse bezüglich der neuronalen Mechanismen beider Reizklassen in einem Übersichtsartikel diskutiert (Studie 5). Gesichter und Körper nehmen eine Sonderstellung innerhalb der visuellen Verarbeitung ein. Im Gegensatz zu Objekten, deren Verarbeitung auf der Wahrnehmung einzelner Details beruht, spielen bei der Gesichter- und Körperwahrnehmung zudem konfigurale Verarbeitungsprozesse eine zentrale Rolle (Studien 2-4). Gesichter und Körper werden in benachbarten kortikalen Arealen verarbeitet (OFA, FFA, EBA, FBA) und evozieren vergleichbare EKPs (P100, N170; siehe Studie 2). Die Befunde der vorliegenden Arbeit deuten darauf hin, dass die konfigurale Körperverarbeitung 170 ms nach der Reizwahrnehmung stattfindet (Studie 3) und dass holistische Verarbeitungsprozesse an der Körperwahrnehmung beteiligt sind (Studie 4). Neben diesen Gemeinsamkeiten wurden auch Unterschiede in den Verarbeitungsmechanismen beobachtet, die beiden Reizkategorien zugrunde liegen (Studie 5): Gesichter und Körper werden im extrastriaten Kortex in distinkten Arealen verarbeitet (gesichtssensitive OFA und körpersensitive EBA). Zudem unterscheidet sich die Quelle der gesichtsspezifischen N170 von der Quelle der körpersensitiven N170 bzw. N190 und holistische Verarbeitungsprozesse scheinen bei der Gesichterverarbeitung, verglichen mit der Körperverarbeitung, eine zentralere Bedeutung zu haben. Das Fehlen einer Körperformagnosie analog der Prosopagnosie stellt zum heutigen Zeitpunkt den größten Unterschied zwischen beiden Kategorien dar.

LITERATUR I FÜNF SEITE 126

5.

LITERATURVERZEICHNIS

Avidan G, Hasson U, Malach R, Behrmann M. Detailed exploration of face-related processing in congenital prosopagnosia: 2. Functional neuroimaging findings. J. Cogn. Neurosci. 2005; 17: 1150-1167. Barton JJ, Cherkasova MV, Press DZ, Intriligator JM, O’Connor M. Developmental prosopagnosia: a study of three patients. Brain. Cogn. 2003; 51: 12-30. Barton JJ, Press DZ, Keenan JP, O’Connor M. Lesions of the fusiform face area impair perception of facial configuration in prosopagnosia. Neurology 2002; 58: 71-78. Behrmann M, Avidan G. Congenital prosopagnosia: face-blind from birth. Trends Cogn. Sci. 2005; 9: 180-187. Behrmann M, Avidan G, Marotta JJ, Kimchi R. Detailed exploration of face-related processing in congenital prosopagnosia: 1. Behavioral findings. J. Cogn. Neurosci. 2005; 17: 1130-1149. Bentin S, Allison T, Puce A, Perez E, McCarthy G. Electrophysiological studies of face perception in humans. J. Cogn. Neurosci. 1996; 8: 551-565. Bentin S, Degutis JM, D’Esposito M, Robertson LC. Too many trees to see the forest: performance, eventrelated potential, and functional magnetic resonance imaging manifestations of integrative congenital prosopagnosia. J. Cogn. Neurosci. 2007; 19: 132-146. Bentin S, Deouell LY, Soroker N. Selective visual streaming in face recognition: evidence from developmental prosopagnosia. Neuroreport 1999; 10: 823-827. Bodamer J. Die Prosopagnosie. Archiv für Psychiatrie und Nervenkrankheiten 1947; 179: 6-53. Collishaw SM, Hole GJ. Featural and configurational processes in the recognition of faces of different familiarity. Perception 2000; 29: 893-909. Cox D, Meyers E, Sinha P. Contextually evoked object-specific responses in human visual cortex. Science 2004; 304: 115-117. Downing PE, Chan AW, Peelen MV, Dodds CM, Kanwisher N. Domain specificity in visual cortex. Cereb. Cortex 2006; 16: 1453-1461.


Downing PE, Jiang Y, Shuman M, Kanwisher N. A cortical area selective for visual processing of the human body. Science 2001; 293: 2470-2473. Downing PE, Wiggett AJ, Peelen MV. Functional magnetic resonance imaging investigation of overlapping lateral occipitotemporal activations using multi-voxel pattern analysis. J. Neurosci. 2007; 27: 226-233. Eimer M. Effects of face inversion on the structural encoding and recognition of faces. Evidence from event-related brain potentials. Brain Res. Cogn. Brain Res. 2000a; 10: 145-158. Eimer M. Event-related brain potentials distinguish processing stages involved in face perception and recognition. Clin. Neurophysiol. 2000b; 111: 694-705. Fairhall SL, Ishai A. Effective Connectivity within the distributed cortical network for face perception. Cereb. Cortex 2007; 17: 2400-2406. Farah MJ, Tanaka JW, Drain HM. What causes the face inversion effect? J. Exp. Psychol. Hum. Percept. Perform. 1995; 21: 628-634. Freire A, Lee K, Symons LA. The face-inversion effect as a deficit in the encoding of configural information: direct evidence. Perception 2000; 29: 159-170. Friston KJ, Buechel C, Fink GR, Morris J, Rolls E, Dolan RJ. Psychophysiological and modulatory interactions in neuroimaging. Neuroimage 1997; 6: 218-229. Friston KJ, Harrison L, Penny W. Dynamic causal modelling. Neuroimage 2003; 19: 1273-1302. Goffaux V, Rossion B. Faces are „spatial“-holistic face perception is supported by low spatial frequencies. J. Exp. Psychol. Hum. Percept. Perform. 2006; 32: 1023-1039. Goodale MA, Milner AD. Separate visual pathways for perception and action. Trends Neurosci. 1992; 15: 20-25. Grill-Spector K, Knouf N, Kanwisher N. The fusiform face area subserves face perception, not generic within-category identification. Nat. Neurosci. 2004; 7: 555-562. Hadjikhani N, de Gelder B. Neural basis of prosopagnosia: an fMRI study. Hum. Brain Mapp. 2002; 16: 176-182.


Harris AM, Duchaine BC, Nakayama K. Normal and abnormal face selectivity of the M170 response in developmental prosopagnosics. Neuropsychologia 2005; 43: 2125-2136. Hasson U, Avidan G, Deouell LY, Bentin S, Malach R. Face-selective activation in a congenital prosopagnosic subject. J. Cogn. Neurosci. 2003; 15: 419-431. Haxby JV, Gobbini MI, Furey ML, Ishai A, Schouten JL, Pietrini P. Distributed and overlapping representations of faces and objects in ventral temporal cortex. Science 2001; 293: 2425-2430. Haxby JV, Hoffman EA, Gobbini MI. The distributed human neural system for face perception. Trends Cogn. Sci. 2000; 4: 223-233. Haxby JV, Ungerleider LG, Clark VP, Schouten JL, Hoffman EA, Martin A. The effect of face inversion on activity in human neural systems for face and object perception. Neuron 1999; 22: 189-199. Herrmann MJ, Ehlis AC, Muehlberger A, Fallgatter AJ. Source localization of early stages of face processing. Brain Topogr. 2005; 18: 77-85. Hole GJ, George PA, Dunsmore V. Evidence for holistic processing of faces viewed as photographic negatives. Perception 1999; 28: 341-359. Ishai A. Let‘s face it: it‘s a cortical network. Neuroimage 2008; 40: 415-419. Ishai A, Ungerleider LG, Martin A, Haxby JV. The representation of objects in the human occipital and temporal cortex. J. Cogn. Neurosci. 2000; 12 Suppl 2: 35-51. Ishai A, Ungerleider LG, Martin A, Schouten JL, Haxby JV. Distributed representation of objects in the human ventral visual pathway. Proc. Natl. Acad. Sci. U.S.A 1999; 96: 9379-9384. Itier RJ, Taylor MJ. Effects of repetition learning on upright, inverted and contrast-reversed face processing using ERPs. Neuroimage 2004a; 21: 1518-1532. Itier RJ, Taylor MJ. Face recognition memory and configural processing: a developmental ERP study using upright, inverted, and contrast-reversed faces. J. Cogn. Neurosci. 2004b; 16: 487-502. Itier RJ, Taylor MJ. N170 or N1? Spatiotemporal differences between object and face processing using ERPs. Cereb. Cortex 2004b; 14: 132-142.


Jacques C, Rossion B. Early electrophysiological responses to multiple face orientations correlate with individual discrimination performance in humans. Neuroimage 2007; 36: 863-873. Kanwisher N, McDermott J, Chun MM. The fusiform face area: a module in human extrastriate cortex specialized for face perception. J. Neurosci. 1997; 17: 4302-4311. Kanwisher N, Tong F, Nakayama K. The effect of face inversion on the human fusiform face area. Cognition 1998; 68: B1-11. Kanwisher N, Yovel G. The fusiform face area: a cortical region specialized for the perception of faces. Philos. Trans. R. Soc. Lond. B Biol. Sci. 2006; 361: 2109-2128. Kovacs G, Zimmer M, Banko E, Harza I, Antal A, Vidnyanszky Z. Electrophysiological correlates of visual adaptation to faces and body parts in humans. Cereb. Cortex 2006; 16: 742-753. Le Grand R, Cooper PA, Mondloch CJ, Lewis TJ, Sagiv N, de Gelder B, Maurer D. What aspects of face processing are impaired in developmental prosopagnosia? Brain Cogn. 2006; 61: 139-158. Le Grand R, Mondloch CJ, Maurer D, Brent HP. Impairment in holistic face processing following early visual deprivation. Psychol. Sci. 2004; 15: 762-768. Leder H, Bruce V. When inverted faces are recognized: the role of configural information in face recognition. Q. J. Exp. Psychol. A 2000; 53: 513-536. Leder H, Candrian G, Huber O, Bruce V. Configural features in the context of upright and inverted faces. Perception 2001; 30: 73-83. Marzi T, Viggiano MP. Interplay between familiarity and orientation in face processing: An ERP study. Int. J. Psychophysiol. 2007; 65: 182-192. Maurer D, Grand RL, Mondloch CJ. The many faces of configural processing. Trends Cogn. Sci. 2002; 6: 255-260. Minnebusch DA, Suchan B, Ramon M, Daum I. Event-related potentials reflect heterogeneity of developmental prosopagnosia. Eur. J. Neurosci. 2007; 25: 2234-2247. Minnebusch DA, Suchan B, Daum I. Losing your head: behavioral and electrophysiological effects of gradual body inversion. J. Cogn. Neurosci. 2008; in press.


Minnebusch DA, Suchan B, Koster O, Daum I. A bilateral occipitotemporal network mediates face perception. Behav. Brain Res. 2009a; 198: 179-185. Minnebusch DA, Daum I. Are human body forms processed hollistically? Evidence from the body composite effect. Cortex; 2009a; submitted. Minnebusch DA, Daum I. Neuropsychological mechanisms of visual face and body perception. Neurosci. Biobehav. Rev., 2009b; submitted Minnebsch DA, Suchan B, Keune PM, Daum I. Gradual inversion effects indicates configural processing of human body shapes. 2009b; in preparation. Moro V, Urgesi C, Pernigo S, Lanteri P, Pazzaglia M, Aglioti SM. The neural basis of body form and body action agnosia. Neuron 2008; 60: 235-246. Morris JP, Pelphrey KA, McCarthy G. Occipitotemporal activation evoked by the perception of human bodies is modulated by the presence or absence of the face. Neuropsychologia 2006; 44: 1919-1927. Norman KA, Polyn SM, Detre GJ, Haxby JV. Beyond mind-reading: multi-voxel pattern analysis of fMRI data. Trends Cogn. Sci. 2006; 10: 424-430. Peelen MV, Downing PE. Selectivity for the human body in the fusiform gyrus. J. Neurophysiol. 2005; 93: 603-608. Peelen MV, Downing PE. The neural basis of visual body perception. Nat. Rev. Neurosci. 2007a; 8: 636648. Peelen MV, Downing PE. Using multi-voxel pattern analysis of fMRI data to interpret overlapping functional activations. Trends Cogn. Sci. 2007b; 11: 4-5. Peelen MV, Wiggett AJ, Downing PE. Patterns of fMRI activity dissociate overlapping functional brain areas that respond to biological motion. Neuron 2006; 49: 815-822.


Pizzagalli DA, Lehmann D, Hendrick AM, Regard M, Pascual-Marqui RD, Davidson RJ. Affective judgments of faces modulate early activity (approximately 160 ms) within the fusiform gyri. Neuroimage 2002; 16: 663-677. Reed CL, Stone VE, Bozova S, Tanaka J. The body-inversion effect. Psychol. Sci. 2003; 14: 302-308. Reed CL, Stone VE, Grubb JD, McGoldrick JE. Turning configural processing upside down: part and whole body postures. J. Exp. Psychol. Hum. Percept. Perform. 2006; 32: 73-87. Righart R, de Gelder B. Impaired face and body perception in developmental prosopagnosia. Proc. Natl. Acad. Sci. U.S.A 2007; 104: 17234-17238. Rossion B. Constraining the cortical face network by neuroimaging studies of acquired prosopagnosia. Neuroimage 2008; 40: 423-426. Rossion B, Caldara R, Seghier M, Schuller AM, Lazeyras F, Mayer E. A network of occipito-temporal face-sensitive areas besides the right middle fusiform gyrus is necessary for normal face processing. Brain 2003; 126: 2381-2395. Rossion B, Delvenne JF, Debatisse D, Goffaux V, Bruyer R, Crommelinck, Guerit JM. Spatio-temporal localization of the face inversion effect: an event-related potentials study. Biol. Psychol. 1999; 50: 173189. Rossion B, Dricot L, Devolder A, Bodart JM, Crommelinck M, de Gelder B, Zoontjes R. Hemispheric asymmetries for whole-based and part-based face processing in the human fusiform gyrus. J. Cogn. Neurosci. 2000a; 12: 793-802. Rossion B, Gauthier I. How does the brain process upright and inverted faces? Behav. Cogn. Neurosci. Rev. 2002; 1: 63-75. Rossion B, Gauthier I, Tarr MJ, Despland P, Bruyer R, Limotte S, Crommelinck M. The N170 occipitotemporal component is delayed and enhanced to inverted faces but not to inverted objects: an electrophysiological account of face-specific processes in the human brain. Neuroreport 2000b; 11: 69-74. Schwarzlose RF, Baker CI, Kanwisher N. Separate face and body selectivity on the fusiform gyrus. J. Neurosci. 2005; 25: 11055-11059. Schweinberger SR, Pickering EC, Jentzsch I, Burton AM, Kaufmann JM. Event-related brain potential


evidence for a response of inferior temporal cortex to familiar face repetitions. Brain Res. Cogn. Brain Res. 2002; 14: 398-409. Sorger B, Goebel R, Schiltz C, Rossion B. Understanding the functional neuroanatomy of acquired prosopagnosia. Neuroimage 2007; 35: 836-852. Steeves JK, Culham JC, Duchaine BC, Pratesi CC, Valyear KF, Schindler I, Humphrey GK, Milner AD, Goodale MA. The fusiform face area is not sufficient for face recognition: evidence from a patient with dense prosopagnosia and no occipital face area. Neuropsychologia 2006; 44: 594-609. Stekelenburg JJ, de Gelder B. The neural correlates of perceiving human bodies: an ERP study on the body-inversion effect. Neuroreport 2004; 15: 777-780. Tanaka JW, Curran T, Porterfield AL, Collins D. Activation of preexisting and acquired face representations: the N250 event-related potential as an index of face familiarity. J. Cogn. Neurosci. 2006; 18: 1488-1497. Taylor JC, Wiggett AJ, Downing PE. fMRI analysis of body and body part representations in the extrastriate and fusiform body areas. J. Neurophysiol. 2007. Thierry G, Pegna AJ, Dodds C, Roberts M, Basan S, Downing P. An event-related potential component sensitive to images of the human body. Neuroimage 2006; 32: 871-879. Ungerleider LG, Haxby JV. ‘What’ and ‘where’ in the human brain. Curr. Opin. Neurobiol. 1994; 4: 157165. Urgesi C, Berlucchi G, Aglioti SM. Magnetic stimulation of extrastriate body area impairs visual processing of nonfacial body parts. Curr. Biol. 2004; 14: 2130-2134. Urgesi C, Calvo-Merino B, Haggard P, Aglioti SM. Transcranial magnetic stimulation reveals two cortical pathways for visual body processing. J. Neurosci. 2007; 27: 8023-8030. Van den Stock J, van de Riet WA, Righart R, de Gelder B. Neural correlates of perceiving emotional faces and bodies in developmental prosopagnosia: an event-related fMRI-study. PLoS.ONE. 2008; 3: e3195. Wada Y, Yamamoto T. Selective impairment of facial recognition due to a haematoma restricted to the right fusiform and lateral occipital region. J. Neurol. Neurosurg. Psychiatry 2001; 71: 254-257. Yin RK. Looking at upside-down faces. J. Exp. Psychol. 1969; 81: 141-145.


Yin RK. Face recognition by brain-injured patients: a dissociable ability? Neuropsychologia 1970; 8: 395-402. Young AW, Hellawell D, Hay DC. Configurational information in face perception. Perception 1987; 16: 747-759. Yovel G, Kanwisher N. Face perception: domain specific, not process specific. Neuron 2004; 44: 889-898. Yovel G, Kanwisher N. The neural basis of the behavioral face-inversion effect. Curr. Biol. 2005; 15: 2256-2262.

Curriculum Vitae

Denise A. Soria Bauser geborene Minnebusch

Am Bugapark 156 – 45899 Gelsenkirchen [email protected]

Persönliche Daten Geboren am: 11.07.1980 in Gelsenkirchen Familienstand: verheiratet

Berufliche Tätigkeiten seit 01/2006

Wissenschaftliche Mitarbeiterin am Lehrstuhl für Neuropsychologie der Ruhr-Universität Bochum Forschungsschwerpunkt: Neuronale Mechanismen der Gesichterund Körperwahrnehmung

Schule und Studium seit 06/2007

Mitglied in der Research School der Ruhr Universität Bochum

seit 01/2006

Doktorandin am Lehrstuhl für Neuropsychologie der RuhrUniversität Bochum

Psychologiestudium, Ruhr-Universität Bochum 12/2004 – 05/2005

Diplomarbeit an der Ruhr-Universität Bochum unter Betreuung von Prof. Dr. I. Daum Thema: Kategoriespezifische Verarbeitungsdefizite bei Prosopagnostikern unter Anwendung des Elektroenzephalogramms (Note: Sehr gut)

05/2003 – 09/2005 Hauptstudium (Note: sehr gut) Diplomprüfungen in den folgenden Fächern: Neuropsychologie, Biopsychologie, Neuroanatomie, Psychologische Diagnostik, Intervention und Evaluation, Psychopathologie

10/2000 – 04/2003

Grundstudium (Note: gut) bestehend aus Methodenlehre, Entwicklungspsychologie, Sozialpsychologie, Testkonstruktion, Allgemeine Psychologie I und II, Physiologie und Differentielle Psychologie und Persönlichkeitsdiagnostik

Schule 08/1991 – 06/2000

Bischöfliches Gymnasium „Am Stoppenberg“, Essen Abitur 06/2000 (Note: 1,6)

Publikationen •

• •

•

Minnebusch DA, Suchan B, Ramon M & Daum I. (2007). Event-related potentials reflect heterogeneity of developmental prosopagnosia. European Journal of Neuroscience, 25: 2234 – 2247. Minnebusch DA, Suchan B, Köster O, Daum I. (2009). A bilateral occipitotemporal network mediates face perception. Behavioural Brain Research, 189: 179-185. Minnebusch DA, Suchan B, Daum I. Losing your head: behavioral and electrophysiological effects of body inversion. Journal of Cognitive Neuroscience (in press). Wolf OT, Minnebusch DA, Daum I. (2009). Stress impairs acquisition of delay eyeblink conditioning in men and women. Neurobiology of Learning and Memory, 91: 431-436.