Neuroradiology (2002) 44: 467–474 DOI 10.1007/s00234-002-0782-2
J. Spreer S. Arnold A. Quiske R. Wohlfarth S. Ziyeh D. Altenmu¨ller M. Herpers J. Kassubek J. Klisch B.J. Steinhoff J. Honegger A. Schulze-Bonhage M. Schumacher Received: 19 June 2001 Accepted: 30 January 2002 Published online: 8 May 2002 Springer-Verlag 2002 J. Spreer (&) Æ S. Arnold Æ S. Ziyeh J. Klisch Æ M. Schumacher Section of Neuroradiology, Neurozentrum, University of Freiburg, Breisgauer Strasse 64, 79106 Freiburg, Germany E-mail:
[email protected] Tel.: +49-761-2705181 Fax: +49-761-2705195 A. Quiske Æ D. Altenmu¨ller A. Schulze-Bonhage Section for Presurgical Epilepsy Diagnosis, Neurozentrum, University of Freiburg, Breisgauer Strasse 64, 79106 Freiburg, Germany R. Wohlfarth Æ B.J. Steinhoff Epilepsiezentrum, Landstrasse 1, 77694 Kehl-Kork, Germany M. Herpers Æ J. Kassubek Department of Neurology, Neurozentrum, University of Freiburg, Breisgauer Strasse 64, 79106 Freiburg, Germany J. Honegger Department of Neurosurgery, Neurozentrum, University of Freiburg, Breisgauer Strasse 64, 79106 Freiburg, Germany
DIAGNOSTIC NEURORADIOLOGY
Determination of hemisphere dominance for language: comparison of frontal and temporal fMRI activation with intracarotid amytal testing
Abstract The reliability of frontal and temporal fMRI activations for the determination of hemisphere language dominance was evaluated in comparison with intracarotid amytal testing (IAT). Twenty-two patients were studied by IAT (bilateral in 13, unilateral in 9 patients) and fMRI using a paradigm requiring semantic decisions. Global and regional (frontal and temporoparietal) lateralisation indices (LI) were calculated from the number of activated (r>0.4) voxels in both hemispheres. Frontolateral activations associated with the language task were seen in all patients, temporoparietal activations in 20 of 22. Regional LI corresponded better with IAT results than global LI. Frontolateral LI were consistent with IAT in all patients with bilateral IAT (including three patients with right dominant and one patient with bilateral language representation) and were not conflicting in any of the patients with unilateral IAT. Temporoparietal LI were discordant
Introduction Determination of hemisphere dominance for language is an important step in diagnostic work-up prior to epilepsy surgery in patients with medically-intractable seizures. Cerebral language representation is highly variable in this patient group [1, 2, 3, 4]. The intracarotid amytal test (IAT) [5] and cortical stimulation [6] remain the ‘‘gold standards’’ in the
with IAT in two patients with atypical language representation. In the determination of hemisphere dominance for language, regional analysis of fMRI activation is superior to global analysis. In cases with clearcut fMRI lateralisation, i.e. consistent lateralised activation of frontal and temporoparietal language zones, IAT may be unnecessary. FMRI should be performed prior to IAT in all patients going to be operated in brain regions potentially involved in language. Keywords Functional magnetic resonance imaging Æ Language Æ Intracarotid amytal test Æ Hemisphere dominance Æ Epilepsy surgery
determination of cerebral language representation. However, the IAT has several disadvantages: it is invasive, time-consuming, and expensive. Due to the short duration of the amytal effect, the number of neuropsychological tests which can be given during the procedure is restricted. IAT results may be obscured by anomalies of the cerebral vasculature or unspecific sedation effects. Cortical stimulation, either direct at surgery or with temporary subdural electrodes, is generally accepted as
468
the most reliable procedure for delineating the extent of ‘‘language-relevant’’ regions [6]. However, the procedures are highly invasive, and the cortical surface which can be examined is restricted to the size of the craniotomy. Functional MRI (fMRI) allows noninvasive assessment of local changes in oxygenation of haemoglobin induced by the bioelectrical activity of neurones [7, 8]. Several reports have shown good correspondence between the results of fMRI language studies and IAT or cortical stimulation [9, 10, 11, 12, 13, 14, 15, 16]. However, there is no generally accepted standard for fMRI language paradigms, data evaluation, and quantitative determination of hemisphere dominance. Most studies have focussed on frontal activation. Several have used verbal fluency or covert word generation paradigms, which have been shown to elicit strongly lateralised frontal activation, but inconsistent activation of temporoparietal areas known to be involved in language. Desmond et al. [9] used a semantic decision paradigm but, because of technical limitations, could examine only one coronal frontal section. Binder et al. [10, 17] suggested another semantic monitoring task, showing good correspondence between IAT and fMRI lateralisation indices (LI). However, LI were calculated from the numbers of suprathreshold voxels in the two hemispheres, ignoring regional differences. Lehe´recy et al. [15] compared IAT with regional (frontal and temporal) LI obtained with different fMRI language paradigms (semantic verbal fluency, covert sentence repetition, story listening). Frontal, but not temporal LI corresponded with IAT. Similar results were obtained by Bahn et al. [12] with verbal fluency and rhyming tasks. Receptive and productive language functions are generally represented in the same cerebral hemisphere. However, interhemispheric dissociation of frontal and temporal language areas was found in 3% of 144 patients with epilepsy undergoing bilateral IAT [18]. Both the frontolateral and temporoparietal language areas should therefore be activated in presurgical fMRI language studies. Lateralisation should be determined separately for both regions. We examined hemisphere dominance for language as determined by IAT and fMRI, using a semantic decision paradigm in 27 patients with medically intractable epilepsy. Our aim was to evaluate the reliability of lateralisation of frontal and temporoparietal activation elicited by the language task.
Materials and methods From 1997 to 2001, 33 patients underwent IAT, of whom six were not examined by fMRI because of cognitive deficits which rendered them unable to perform the necessary tasks. The 27 who were examined by fMRI and IAT all gave informed consent. The fMRI studies of five patients could not be assessed because of artefacts: in
one the stimulation computer did not work, and four studies were impaired by severe motion artefact which could not be removed by motion correction. The clinical and pathological data of the remaining 10 women and 12 men are given in Table 1. According to the Edinburgh handedness inventory [19], 15 patients were right-handed, four left-handed and three ambidextrous. There were 20 with medicallyintractable seizures and three with seizures caused by spaceoccupying lesions requiring surgery. All patients were treated with anticonvulsants. In patient 3 no lesion was found on MRI; a seizure focus was in the right temporal lobe. Patient 17 had bilateral Ammons horn sclerosis; the seizure focus was on the right. Patient 20, with tuberous sclerosis, had multiple lesions; the seizures originated from a left frontal tuber. In all other patients the seizures originated from the lesions listed in Table 1. The lesions were benign, except in one patient who had a glioblastoma multiforme, although in seven patients who did not undergo surgery, the histology was simply suspected from MRI. Ammons horn sclerosis, present in 11 patients, was the most common diagnosis. Prior to IAT, we performed transfemoral intra-arterial cerebral angiography to exclude vascular anomalies. During the procedure, the patients were continuously monitored by a neurologist and by video-EEG-telemetry. After injection of sodium amytal in the internal carotid artery and the onset of hemiparesis, we carried out neuropsychological examination of language functions, including counting, body commands, naming, recognition of drawings of concrete objects (modified Token test), sentence repetition and reading. We classified patients as left dominant when there was no language impairment on right-sided injection and marked dysphasia occurred with left-sided injection. Patients with a reversed pattern: no dysphasia on left injection, dysphasia on right IAT, we classified as right dominant. Any patient not belonging to one of these categories we classified as bilateral. Initial impairment of counting may occur with anaesthesia of the nondominant hemisphere [20], so that isolated disturbance of counting without additional language impairment was not taken to indicate dominance for language of the injected hemisphere. In patients 1 to 8, we generally performed the IAT bilaterally, giving 125 mg sodium amytal per side with a 1-h interval between injections. The IAT regime was changed starting with patient 9, when IAT was routinely performed on the side of surgery; the dose of sodium amytal was increased (women: 140 mg; men: 170 mg). The procedure was repeated on the other side after at least 24 h only in patients with ambiguous results concerning expected postoperative language deficits. A unilateral IAT does not exclude positive or negative bilaterality and allows no definite conclusion regarding hemisphere dominance [18]. Therefore, in patients with unilateral IAT a ‘‘presumed dominance’’ is given in Table 2. The fMRI paradigm was developed from a verbal subtest of a German intelligence test, the Wilde Intelligence Test [21], and has previously been shown to elicit lateralised cortical activation in healthy controls [22]. The patient is required to find a pair of synonyms out of a set of five words presented simultaneously (Fig. 1). A parallelised colour-discrimination task (Fig. 2) was chosen as the baseline condition, to compensate for nonspecific effects such as eye movements, primary visual activation and attention [23]. The stimuli were generated on a personal computer (PC), stimulus presentation being triggered by the MR imager. The display was projected onto a screen at the head end of the tunnel of the magnet, which the patient saw in a mirror mounted on the head coil. The patient had to press with the dominant hand one of four keys on a custom-made MRI-compatible box. The responses were recorded on a PC. A fixed study design was used in all patients. The paradigm was arranged as a block design with eight alternating periods of colour discrimination and language tasks, with an additional baseline period at the end. Every period consisted of four
469
Table 1. Clinical data Patient
Age (years), sex
Age (years) at seizure onset
Edinburgh handedness scorea
Pathology
Seizure focus
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
28, 36, 34, 26, 24, 38, 17, 51, 27, 15, 49, 42, 18, 23, 34,
f f f f f m f m f m m f m m m
12 12 2 16 20 5 1 20 8 5 17 2 2 1 32
+88 +100 –86 +100 +100 –88 +7 –8 +91 –84 +100 +84 +100 +86 +50
Left temporal Right temporal Right temporal Right temporal Right temporal Left temporal Left occipital Right frontal Right temporal Right temporal Right frontal Left temporal Right temporal Left frontal Right temporal
16 17 18 19 20 21 22
28, 34, 68, 32, 16, 38, 26,
m f m m m f m
26 6 68 20 6 17 18
+83 +100 +100 +60 +87 +69 –83
Ammon’s horn sclerosis Ganglioglioma No lesion Ammon’s horn sclerosis Ammon’s horn sclerosis Ammon’s horn sclerosis Dysplasia Cavernoma Ammon’s horn sclerosis Ammon’s horn sclerosis Abscess Ammon’s horn sclerosis Ammon’s horn sclerosis Cortical dysplasia Dysembryoplastic neuroectodermal tumour Astrocytoma II Bilateral Ammon’s horn sclerosis Glioblastoma Contusion Tuberous sclerosis Cavernomas Dysplasia
a
Right frontal Right temporal Right temporal Left frontal Left frontal Left temporal Right parietooccipital
positive values: right hand preference
Table 2. Hemisphere dominance for language as determined by intracarotid amytal testing and fMRI. See text for derivation of fMRI dominance
Patient
IAT dominance
Patients with bilateral IAT 1 Left 2 Left 3 Right 4 Left 5 Left 6 Left 7 Ambiguous 8 Bilateral 9 Right 12 Left 17 Left 19 Right 20 Left Patients with unilateral IAT Patient IAT side Presumed IAT dominance 10 Right Left 11 Right Left 13 Right Left 14 Right Left 15 Right Left 16 Right Left 18 Right Right 21 Left Left 22 Right Left
tasks, each was projected for 5 s. Prior to the examination, the patients were instructed outside the imaging room, using test versions of the tasks.
FMRI dominance Global
Frontolateral
Temporoparietal
Bilateral Left Bilateral Left Left Left Left Bilateral Right Bilateral Left Right Left
Left Left Right Left Left Left Left Bilateral Right Left Left Right Left
Left Left Left Left Left Left Bilateral Left Right Left Left Right Left
Global
Frontolateral
Temporoparietal
Left Left Left Left Left Left Bilateral Left Bilateral
Bilateral Left Left Left Left Left Bilateral Left Bilateral
Left Left Left No activation Left Left No activation Left Left
The examinations were performed on a 1.5 tesla imager with a standard circularly polarised head coil. Multislice gradient-echo echoplanar imaging (EPI) sequences were used for the functional
470
the number of suprathreshold voxels within the different ROI. LI+0.2 as right dominant [10].
Results
Fig. 1. Language task: the patient has to find the synonym to the uppermost word out of a set of words presented simultaneously
FroLat fMRI language-task activation was observed in all 22 and TemPar activation in 20 of the 22 studies. A histogram of the distribution of the LI is given in Fig. 3. No significant correlation was found between performance data (error quotes and reaction times) and fMRI activation. The IAT and the fMRI results are summarised in Table 2. FroLat LI showed the best agreement with the IAT and were concordant in all patients with bilateral IAT (including four with atypical, i.e., right-hemisphere or bilateral language representation) (Table 2). The global LI yielded bilateral language representation in three patients with lateralised IAT results. TemPar LI were discordant with IAT in two patients (Figs. 4, 5, 6). fMRI and IAT did not yield conflicting data in any patient with unilateral IAT (Table 2b). Since IAT was performed bilaterally in only 13 patients, we did not carry out a statistical analysis. Chi-square analysis on a dichotomised level (concordance yes/no) would have obscured the misleading lateralisation by the TemPar LI in two patients. IAT-fMRI correlation did not depend on the type of pathology. We carried out surgery on 17 patients. In one patient predicted to be left-dominant by IAT and fMRI the left middle cerebral artery was injured at surgery, resulting in intracerebral haemorrhage with concomitant aphasia. No other patient developed a language deficit.
Fig. 2. Colour discrimination task: the patient has to find the colour corresponding to the uppermost colour field
Discussion studies. We acquired 72 volumes consisting of 16 axial slices (5 mm thick, gap 0.5 mm, field of view 256 mm, TR 5 s). The matrix was 128·128 (TE 84 ms) for the first eight patients, and after that 64·64 (TE 64 ms). To compensate for the initial signal decay before nett magnetisation equilibrium, the first four scans of every functional series were discarded. Following the functional studies, we acquired anatomical T1-weighted spin-echo images (TR 450 TE 14 ms, field of view 256 mm, matrix 256·256) in the same plane of section. We analysed the data with BrainVoyager software [24], applying motion correction, spatial and temporal filtering and eliminating linear drifts. We performed a correlation analysis with a boxcar reference. Voxels with ‘‘significant’’ (r>0.4) signal increase in the language tasks were colour-coded and superimposed on the anatomical T1-weighted images. To quantify hemisphere dominance, LI were calculated from the number of activated voxels [10]: LI=[Svoxel right–Svoxel left]/ Svoxel right+left; LI=–1 indicated exclusively left, and LI=+1 exclusively right hemisphere activation). LI were calculated for the cerebral hemispheres as a whole (global, excluding infratentorial regions), and for two regions of interest (ROI): frontolateral convexity: inferior and medial frontal gyri, and the convexity of the superior frontal gyrus (FroLat); and the superior and middle temporal, supramarginal and angular gyri (TemPar). The ROI were defined by visual inspection in every individual data-set. BrainVoyager allows semiautomatic counting of
In the last decade functional neuroimaging has been used extensively to study cerebral representation of
Fig. 3. Histogram of global and frontolateral (FroLat) lateralisation indices (LI). Note almost continuous distribution of LI with predominance of leftward asymmetry. Frontolateral LI ‘‘lateralise’’ stronger than global LI
471
Fig. 4. Exclusive left hemisphere activation of frontolateral and temporoparietal language zones in an ambidextrous patient; colour-coding of ‘‘activated’’ (r>0.4) pixels; yellow indicating higher
Fig. 5. Right hemisphere language dominance in a right-handed patient, in accordance with intracarotid amytal testing
language [25, 26]. In accordance with several previous studies [9, 10, 11, 15, 17, 25] frontal activation elicited by our language paradigm generally extended over the
Fig. 6. Predominant activation of the right frontolateral and left temporoparietal language zones. In intracarotid amytal testing, only injection on the right produced aphasia
frontal operculum, i.e., the ‘‘classical’’ Broca’s area, and included the middle and/or superior frontal gyri. This extended frontal activation reflects different cognitive processes involved in the semantic decision task, such as semantic analysis [17, 23, 27, 28, 29, 30], semantic encoding [31], phonological processing [27] and/or storage and rehearsal in verbal working memory [32]. The temporoparietal language zone includes several multimodal areas responding to auditory and visual verbal stimuli [25]. It is involved in phonologic analysis [33] and may participate in lexical (graphemic) processing [30, 34, 35]. With our semantic decision task we found temporoparietal activation in 20 of 22 studies. The choice of an appropriate baseline condition is crucial to cognitive fMRI. Our colour discrimination task was adapted from work by Spitzer et al. [23], who showed that active control conditions give better fMRI language studies than comparisons between language tasks and resting conditions. It might be argued that the baseline condition we used induces additional colour processing, a cerebral process not involved in language. However, the cortical sites mainly involved in colour processing are in the fusiform and lingual gyri, clearly distinct from the classical frontolateral and temporoparietal ‘‘language areas’’. We cannot exclude the possibility that implicit naming of the colours could have occurred in some patients during the colour discrimination task, thus reducing apparent activation, as this is derived from a statistical comparison of the two conditions. This problem might be diminished with other
472
active baseline tasks such as visual judgements on pseudofonts or letter strings. Since productive and receptive language functions in patients with epilepsy are not necessarily served by the same hemisphere [18], we have indicated that the frontolateral and temporoparietal language areas should be activated in presurgical fMRI language studies. In several previous studies, verbal fluency or word generation paradigms have been used, which reliably activate frontolateral language areas, but inconsistently activate of the temporoparietal language zone [12, 15]. As in the studies by Lehe´recy et al. [15] and Bahn et al. [12], Frontolateral LI gave better agreement with IAT than global LI. Using global LI, we classified three patients as ‘‘bilateral’’, whereas frontolateral LI and IAT yielded lateralised results. These findings stress the importance of regional analysis in fMRI language studies. Global LI, as suggested by several groups [10, 14, 36], may be misleading. However, temporoparietal LI were discordant with IAT in two patients, both of whom had atypical language representation (one right-dominant and one bilateral). Although the IAT generally is accepted as the ‘‘gold-standard’’, it may yield false results in patients with atypical language representation. Wyllie et al. [3] examined 88 patients by bilateral IAT and cortical stimulation of the hemisphere predicted not to be involved in language. In two of nine classified as rightdominant by IAT, cortical stimulation of lefthemisphere areas induced language disturbances. It might be speculated that our patients with discordant IAT and fMRI had similar language representation. However, we assume that, in cases with discordant lateralisation in frontal and temporoparietal language areas, fMRI should be confirmed by deactivation methods (i.e., IAT or electrocorticography). Since ‘‘brain activation’’ in fMRI reflects local haemodynamic changes associated with electrical neuronal activity, it depends on intact neurovascular coupling mechanisms. Pathological processes influencing neurovascular coupling may alter the blood oxygen level-dependent (BOLD) signal and the results of brain activation studies. Reduced BOLD signal in response to motor tasks in cortical areas infiltrated by gliomas has been shown by Schreiber et al. [37] and Holodny et al. [38]. On the other hand, Ojemann et al. [39] reported a patient with an anaplastic astrocytoma of the supplementary motor area, in whom language tasks evoked an disproportionately large increase in regional cerebral blood flow in the region invaded by the tumour. In one study [14], the results of fMRI and IAT were discordant in one patient with a huge left-hemisphere tumour, but became concordant when the tumour area and the right hemisphere analogues were excluded from calculation of lateralisation.
None of our patients had a large tumour invading frontal or temporoparietal regions potentially involved in language. However, the results of fMRI language studies have to be interpreted with caution in such cases. The same holds true for patients with vascular malformations, since the BOLD effect may be influenced in an unpredictable manner. Stenoses of arteries supplying brain may reduce the cerebrovascular reserve capacity and decrease task-associated changes in blood flow in brain activation studies, falsifying LI [40]. According to clinical conventions, cerebral language representation is commonly divided into distinct categories, e.g. left-dominant, bilateral, and right-dominant. However, the LI in the patients in the present study were almost continuously distributed (with a strong preference of the left hemisphere) and not discrete (Fig. 3). This is in accordance with several previous studies in both patients and healthy subjects [10, 14, 36]. LI calculated from increases in blood flow in the middle cerebral arteries induced by language tasks also show a continuous distribution [41]. If IAT results are quantified separately for productive and receptive language functions in both hemispheres, the LI are similar to those obtained in brain activation studies. Binder et al. [10] and Benson et al. [14] showed good correspondence between LI obtained with IAT and fMRI (although in both studies different tasks were applied in IAT and fMRI). Thus, the concept of three categories of cerebral language representation (left, right, bilateral) is obviously a simplification. On the other hand, from the surgical point of view, a clear decision is required as to whether a surgical procedure may interfere with language functions or not. An appropriate cut-off for the definition of language dominance is therefore crucial. We chose LI of +/–0.2, as suggested by Binder et al. [10]. However, with higher cut-offs, the percentage of patients classified as bilateral will decrease and the percentage of ‘‘lateralised’’ patients increase. Can fMRI replace IAT? They are complementary: fMRI is an activation method, IAT is a deactivation procedure. Activation studies show brain regions participating in certain tasks, whereas deactivation studies reveal which regions are essential to achieve the tasks. As stated above, the IAT may yield false results in patients with atypical language representation [3]. Discrepancies between fMRI and IAT are therefore not necessarily shortcomings of fMRI. Nevertheless, at this time IAT is considered to be the least ambiguous method in the determination of hemisphere dominance. At present, fMRI certainly cannot replace the IAT in all patients. In the interpretation of fMRI studies, the results of other noninvasive (neuropsychological tests) and invasive procedures have to be taken into account. However, we think that fMRI should generally be performed prior to IAT because it may provide useful
473
additional preoperative information in candidates for epilepsy surgery. In addition to determination of hemisphere dominance, fMRI gives information about the localisation of language-relevant areas. In patients with clear-cut dominance, i.e., consistently lateralised activation of frontal and temporoparietal areas known to be
involved in language, and without a pathological process influencing neurovascular coupling, IAT may be avoided. Further development and evaluation of fMRI paradigms is required to reduce the percentage of patients in whom invasive procedures are necessary to determine cerebral language representation.
References 1. Rasmussen T, Milner B (1977) The role of early left-brain injury in determining lateralization of cerebral speech functions. Ann N Y Acad Sci 299:355–369 2. Woods RP, Dodrill CB, Ojemann GA (1988) Brain injury, handedness, and speech lateralization in a series of amobarbital studies. Ann Neurol 23:510–518 3. Wyllie E, Lu¨ders H, Murphy D, et al (1990) Intracarotid amobarbital (Wada) test for language dominance: correlation with results of cortical stimulation. Epilepsia 31:156–161 4. Kurthen M, Helmstaedter C, Linke DB, Hufnagel A, Elger CE, Schramm J (1994) Quantitative and qualitative evaluation of patterns of cerebral language dominance: an amobarbital study. Brain Lang 46:536–564 5. Wada J, Rasmussen T (1960) Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance: experimental and clinical observations. J Neurosurg 17:266–282 6. Ojemann G, Ojemann J, Lettich B, et al (1989) Cortical language localization in left dominant hemisphere. J Neurosurg 71:316–326 7. Belliveau JW, Kennedy DN, McKinstry RC et al (1991) Functional mapping of the human visual cortex by magnetic resonance imaging. Science 254:716– 719 8. Kwong KK, Belliveau JW, Chesler DA, et al (1992) Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. Proc Natl Acad Sci USA 89:5675–5679 9. Desmond JE, Sum JM, Wagner AD, et al (1995) Functional MRI measurement of language lateralization in Wada-tested patients. Brain 118:1411– 1419 10. Binder JR, Swanson SJ, Hammeke TA, et al (1996) Determination of language dominance using functional MRI: a comparison with the Wada test. Neurology 46:978–984 11. Hertz-Pannier L, Gaillard WD, Mott SH, et al (1997) Noninvasive assessment of language dominance in children and adolescents with functional MRI: a preliminary study. Neurology 48:1003– 1012
12. Bahn MM, Lin W, Silbergeld DL, et al (1997) Localization of language cortices by functional MR imaging compared with intracarotid amobarbital hemispheric sedation. Am J Roentgenol 169:575–579 13. Yetkin FZ, Swanson S, Fischer M, et al (1998) Functional MR of frontal lobe activation: comparison with Wada language results. AJNR 19:1095–1098 14. Benson RR, Fitzgerald DB, LeSueur LL, et al (1999) Language dominance determined by whole brain functional MRI in patients with brain lesions. Neurology 52:798–809 15. Lehe´ricy S, Cohen L, Bazin B, et al (2000) Functional MR evaluation of temporal and frontal language dominance compared with the Wada test. Neurology 54:1625–1633 16. Bazin B, Cohen L, Lehe´ricy S, et al (2000) Study of hemispheric lateralization of the language regions by functional MRI. Validation with the Wada test. Rev Neurol 156:145–148 17. Binder JR, Frost JA, Hammeke TA, Cox RW, Rao SM, Prieto T (1997) Human brain language areas identified by functional magnetic resonance imaging. J Neurosci 17:353–362 18. Kurthen M, Helmstaedter C, Linke DB, Solymosi L, Elger CE, Schramm J (1992) Interhemispheric dissociation of expressive and receptive language functions in patients with complexpartial seizures: an amobarbital study. Brain Lang 43:694–712 19. Oldfield RC (1971) The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 9:97–113 20. Benbadis SR, Binder JR, Fischer M, et al (1998) Is speech arrest during Wada testing a valid method for determining hemispheric representation of language? Brain Lang 65:441–446 21. Ja¨ger AO, Althoff K (1994) Der WILDE-Intelligenz-Test. Ein Strukturdiagnostikum. Verlag Psychologie Dr.C.J.Hogrefe, Go¨ttingen 22. Spreer J, Ziyeh S, Wohlfahrt R, et al (1998) Vergleich verschiedener Paradigmen fu¨r die fMRT zur Bestimmung der Hemispha¨rendominanz fu¨r sprachliche Funktionen. Klin Neuroradiol 8:173–181
23. Spitzer M, Bellemann ME, Kammer T, et al (1996) Functional MR imaging of semantic information processing and learning-related effects using psychometrically controlled stimulation paradigms. Cogn Brain Res 4:149–161 24. Goebel R, Khorram-Sefat D, Muckli L, Hacker H, Singer W (1998) The constructive nature of vision: direct evidence from functional magnetic resonance imaging studies of apparent motion and motion imagery. Eur J Neurosci 10:1563–1573 25. Price CJ (2000) The anatomy of language: contributions from functional neuroimaging. J Anat 197:335–359 26. Cabeza R, Nyberg LJ (2000) Imaging cognition II: an empirical review of 275 PET and fMRI studies. Cogn Neurosci 12:1–47 27. Poldrack RA, Wagner AD, Prull MW, Desmond JE, Glover GH, Gabrieli JD (1999) Functional specialization for semantic and phonological processing in the left inferior prefrontal cortex. Neuroimage 10:15–35 28. Demb JB, Desmond JE, Wagner AD, Vaidya CJ, Glover GH, Gabrieli JD (1995) Semantic encoding and retrieval in the left inferior prefrontal cortex: a functional MRI study of task difficulty and process specificity. J Neurosci 15:5870–5878 29. Petersen SE, Fox PT, Snyder AZ, Raichle ME (1990) Activation of extrastriate and frontal cortical areas by visual words and word-like stimuli. Science 249:1041–1044 30. Price CJ, Wise RSE Watson JDG, et al (1994) Brain activity during reading: the effects of exposure duration and task. Brain 117:1255–1269 31. Tulving E, Kapur S, Craik FI, Moscovitch M, Houle S (1994) Hemispheric encoding/retrieval asymmetry in episodic memory: positron emission tomography findings. Proc Natl Acad Sci USA 91:2016–2020 32. Paulesu EM, Frith CD, Frackowiak RSJ (1993) The neural correlates of the verbal component of working memory. Nature 362:342–345
474
33. Wise RJS, Scott SK, Blank SC, Mummery CJ, Murphy K, Warburton EA (2001) Separate neural subsystems within ‘‘Wernicke’s area’’. Brain 124:83–95 34. Frith CD, Friston KJ, Liddle PF, Frackowiak RSJ (1991) A PET study of word finding. Neuropsychologia 29:1137–1148 35. Howard D, Patterson K, Wise R, et al (1992) The cortical localisation of the lexicons: positron emission tomography evidence. Brain 115:1769–1782
36. Springer JA, Binder JR, Hammeke TA, et al (1999) Language dominance in neurologically normal and epilepsy subjects. Brain 122:2033–2045 37. Schreiber A, Hubbe U, Ziyeh S, Hennig J (2000) The influence of gliomas and nonglial space-occupying lesions on blood-oxygen-level-dependent contrast enhancement. AJNR 21:1055–1063 38. Holodny AI, Schulder M, Liu WC, Wolko J, Maldjian JA, Kalnin AJ (2000) The effect of brain tumors on BOLD functional MR imaging activation in the adjacent motor cortex: implications for image-guided neurosurgery. AJNR 21:1415–1422
39. Ojemann JG, Neil JM, MacLeod AM, et al (1998) Increased functional vascular response in the region of a glioma. J Cereb Blood Flow Metab 18:148–153 40. Spreer J, Yahya H, Raab P (2000) Fehlende ha¨modynamische Antwort bei erhaltener neuronaler Funktion: Pitfall der funktionellen MRT bei Patienten mit stenosierenden Gefa¨ßprozessen. Fortschr Geb Rontgenstr 172:398–400 41. Knecht S, Dra¨ger B, Deppe M, et al (2000) Handedness and hemispheric dominance in healthy humans. Brain 123:2512–1251
