Exp Brain Res (2000) 132:163–178 Digital Object Identifier (DOI) 10.1007/s002219900328
R E S E A R C H A RT I C L E
Rob H. J. van der Lubbe · Bernd Wauschkuhn Edmund Wascher · Torsten Niehoff · Detlef Kömpf Rolf Verleger
Lateralized EEG components with direction information for the preparation of saccades versus finger movements Received: 28 October 1999 / Accepted: 21 December 1999 / Published online: 23 February 2000 © Springer-Verlag 2000
Abstract During preparation of horizontal saccades in humans, several lateralized (relative to saccade direction), event-related EEG components occur that have been interpreted as reflecting activity of frontal and parietal eye fields. We investigated to what degree these components are specific to saccade preparation. EEG lateralization was examined within the interval (1 s) between a first (S1) and a second (S2) stimulus, after which a response had to be made (look left or right, or press a button with the left or right index finger). The visual S1 indicated either the direction (left vs right) and/or the effector (eye vs finger), and S2 (visual/auditory in different blocks) added the information not given by S1. An occipital component (220 ms after S1) was effector-independent, probably reflecting processing of the direction code. The following parietotemporal component (320 ms after S1) was specific for direction information. This component seems more relevant for finger movements than for saccades and may reflect a link between visual perception to action. A later frontal component (480 ms after S1) was specific for direction information and may be related to the planning of a lateral movement. One component was entirely specific for the preparation of a finger movement (the lateralized readiness potential before S2). Thus, several different lateralized processes in the S1-S2 interval could be delineated, reflecting hand-specific preparation, processing of the direction code, and the coordination of perception and action, but no components were observed as being specific for saccade preparation.
R.H.J. van der Lubbe (✉) · B. Wauschkuhn · T. Niehoff D. Kömpf · R. Verleger Klinik für Neurologie, Medizinische Universität zu Lübeck, Ratzeburger Allee 160, D-23538 Lübeck, Germany e-mail:
[email protected] Tel.: +49-451-5003544, Fax: +49-451-5002489 E. Wascher Institute of Clinical and Physiological Psychology, University of Tübingen, Tübingen, Germany
Key words Saccades · Response preparation · Event-related lateralizations · Finger movements · Principal component analysis · Human
Introduction In the last decade, many studies have shown that the computation of differences of event-related potentials (ERPs), contralateral minus ipsilateral, to the response or to the stimulus can provide important facts about human information processing, because the lateralized processing is specific to the relevant (lateral) stimulus or to the selected (lateral) response (De Jong et al. 1988; Eimer 1995, 1996, 1997; Gratton et al. 1988, 1997; van der Lubbe and Woestenburg 1997, 1999, 2000; Luck and Hillyard 1994a, 1994b; Osman et al. 1992; Wascher and Wauschkuhn 1996; Wascher et al. 1999). In this study, these ERP differences will be denoted as event-related lateralizations (ERLs). In a study by Wauschkuhn et al. (1997), ERLs were investigated to delineate saccadespecific lateralized preparation. During an interval between a first (S1) and a second stimulus (S2), saccades and finger movements had to be prepared which evoked different ERL components. The main goal of the current study was to replicate and extend these findings in conditions wherein response preparation was assumed to be more optimal and the possibility for interpretation was made less ambiguous. A principal component analysis (PCA) was used to decompose the observed lateralizations which is a more objective and comprehensive method of describing the data than the analysis of selected time intervals used by Wauschkuhn et al. (1997). Special emphasis was placed on a parietal ERL component that might represent the output of a general decoding process of direction information (Mesulam 1990) and on a frontocentral ERL component that might reflect saccade-specific preparation. The study by Wauschkuhn et al. (1997) was inspired by a study of Klostermann et al. (1994) in which participants performed self-paced volitional saccades to the
164
right, aimed either at a position in the dark to be remembered or at a visual target. Negativity before saccade onset was larger over the left hemisphere, contralateral to the direction of the saccade, most marked at the left parietal electrode (P3) in the condition with visual targets. In addition, a temporary negativity maximal at the frontocentral electrode contralateral to the direction of the saccade (FC3) was observed about 1 s before saccade onset. No such effects were found in the dark condition. The frontocentral effect was hypothesized to reflect activity of the frontal eye field (FEF) contralateral to the saccade target (Schall 1991), specific for the preparation of a saccade. The parietal effect was attributed to contralateral activity in the posterior parietal cortex, reflecting the focusing of attention at the visual target. Participants only performed saccades to the right, therefore the observed effects may simply be due to the different involvement of the left and right hemisphere in general activation processes. Furthermore, the interpretation of the parietal and the frontocentral component appears rather premature, as these components may also reflect other processes such as the planning of a lateral movement independent of the required effector, or the frontocentral component may also reflect the focusing of attention (Thickbroom and Mastaglia 1990). Similar arguments can be made with regard to enhancements of event-related activity contralateral to saccade direction reported in other studies. For example, several studies (Evdokimidis et al. 1992; Everling et al. 1998; Wauschkuhn et al. 1998) reported a “saccade-onset related lateralization” 100–50 ms before saccade onset, maximum at P3/P4, although it was not demonstrated that this lateralization was specific to saccades. The S1-S2 partial-cueing study of Wauschkuhn et al. (1997) was designed to clarify these issues. After arrival of S2, either a left or a right saccade to a visual target had to be performed, or a left or right button had to be pressed with the index finger of the corresponding hand. S1 either gave no information or indicated either the direction (left or right) or the effector (eye vs hand) or both the direction and the effector (full information) of the required response. S2 always gave complete information about the required response, irrespective of that information having been given by S1. It should be noted that a consequence may be that the information given by S1 was not decoded completely, because this information was also given as S2. ERLs were computed where contralateral and ipsilateral were defined relative to the required response side.1 Wauschkuhn et al. (1997) distinguished three ERL components: a negative component at 200–400 ms after S1 (largest at P7/8, denoted as “early temporoparietal 1 This is an extension of the method used to compute the lateralized readiness potential (LRP; De Jong et al. 1988; Gratton et al. 1988) to all symmetrical electrode positions (Wascher and Wauschkuhn 1996) and results in subtracting out both nonlateralized potentials and lateralized effects that are not related to the required response side, e.g., the different involvement of the left and right hemisphere in general activation processes.
lateralization,” ETPL); a positive component from 500 ms onwards after S1 (largest at P7/8, denoted as “late temporoparietal lateralization,” LTPL); and the lateralized readiness potential (LRP, largest at C1/2) starting from 750 ms after S1, which can be interpreted as reflecting motor activation (De Jong et al. 1988; Eimer 1995; Gratton et al. 1988). The ETPL was equally large when S1 indicated the required saccade or the required finger movement, whereas it was smaller when S1 only indicated the required response direction. The LRP and LTPL were only observed when S1 indicated the required finger movement. The latter finding indicates that participants did not equally prepare a finger movement when S1 only indicated the required response direction compared with when S1 gave full information about the required finger movement. Unfortunately, no analyses were reported regarding the frontocentral sites, although a frontocentral component may have been present. If we relate the findings of Wauschkuhn et al. (1997) to the observations made by Klostermann et al. (1994), then it appears that both studies provided support for a parietal or temporoparietal ERL component (the ETPL in the study by Wauschkuhn et al., 1997), whereas some support for a frontocentral component was only provided by Klostermann et al. (1994). First, we will discuss some possible interpretations of the parietal component. Thereafter, we will shortly outline a working model for the processes that may take place while preparing for a finger movement or saccade. Klostermann et al. (1994) argued that the parietal component probably reflects the focusing of attention at the visual target (i.e., attentional orienting; see also Van ‘t Ent and Apkarian 1998; Wauschkuhn et al. 1998). This view is supported by studies that also revealed a parietal, lateralized component in the S1-S2 interval, which was interpreted as a reflection of attentional orienting to one side (Harter et al. 1989; Yamaguchi et al. 1994), and by studies that proposed a strict link between orienting of attention and programing of ocular movements (Duhamel et al. 1992; Posner and Petersen 1990; Rizzolatti et al. 1987). For instance, Rizzolatti et al. (1987) suggested that attention is oriented to a given point when the oculomotor program for moving the eyes to this point is ready to be executed. In the study of Wauschkuhn et al. (1997), attentional orienting to one side makes sense only when a saccade has to be performed to the lateral saccade target and not when a finger movement has to be performed. Yet their temporoparietal component was also present when S1 indicated a finger movement, which makes the interpretation of attentional orienting unlikely. The same reasoning can be applied against the interpretation that the component reflects saccade-specific lateralized activity, which may originate from the parietal eye field (Pierrot-Deseilligny et al. 1995), as no component should be found when S1 indicated the required hand movement. It should be noted that arrows may also automatically induce attention or response tendencies independent from the required response or from the relevant
165 Fig. 1 Two examples (middle and right) of the sequence of stimuli in the visual and auditory S2 conditions. An example of a saccade trial in which effector information was given by S1 and direction information was provided by S2 is depicted in the middle panel. An example of a finger movement trial in which direction information was provided by S1 is presented in the right panel. The event durations are indicated in the left column
focus of attention2 (see Eimer 1995, 1997). Thus, the attentional interpretation of this component may be saved by an automatic tendency to attend to the side indicated by the arrow. However, Eimer (1997) found no temporoparietal component when uninformative (50% valid) arrows indicated the relevant stimulus side. Another possibility is that the temporoparietal component is an exogenous effect, due to the small asymmetry of the centrally presented arrows (see Fig. 1). However, such an explanation is discredited by the late latency of the effect (200–400 ms) and by the finding that the component was larger when S1 gave full information about the required response, which may be interpreted as more efficient preparation in the case of full information. In addition, a recent study by van der Lubbe et al. (1999) showed a similar lateralization when letters were used as S1. Thus, the effect seems not due to the small asymmetry of the arrows or to an automatic effect specific for arrows. Wauschkuhn et al. (1997) favored another explanation of their results, which they based on studies by Mesulam (1990) and Rizzolatti and Berti (1990),3 namely that the parietal component “represents the output of a general decoding process of directional information in the posterior parietal cortex which leads to increased activity of the cortex contralateral to the direction of the decoded direction signal. In the next step of movement preparation this information may be transferred to the 2 3
This possibility was suggested by a reviewer of this paper. This view dates back to the work of Mountcastle et al. (1975), who argued that the posterior parietal cortex has a command function for the initiation of motor and oculomotor behavior; see also, more recently, Milner and Goodale 1995.
frontal eye field for the preparation of a saccade and to the motor cortex...for the preparation of a finger movement.” Thus, the temporoparietal component may reflect a supramodal direction code, used for the control both of hand movements and of saccades. As a working model for the processes that may take place while preparing for a required response, we followed the ideas of Wauschkuhn et al. (1997). A distinction was made between two preparatory processes. First, a general decoding process is employed that produces a supramodal direction code on the basis of sensory information. This process may be reflected in the parietal component (the “ETPL” of Wauschkuhn et al., 1997). Second, the output of the general decoding process may be used by effector-specific processes to activate the required response. This process may be reflected in a frontocentral component when a saccade has to be executed (originating from the FEF), and in a central lateralization (the LRP) and a parietal positive component (the “LTPL” of Wauschkuhn et al., 1997) when a finger movement has to be performed. In the current study, two important modifications of the paradigm of Wauschkuhn et al. (1997) were made. First, in our experiment S2 only provided the information not given by S1, whereas, in the study by Wauschkuhn et al. (1997), S2 always gave complete information about the required response, irrespective of that information having been given by S1. This modification ensures that the information given by S1 is processed within the S1-S2 interval, which increases the possibility of finding ERL components reflecting specific preparatory processes. Second, we added a condition
166
in which S2 was auditory instead of visual. We reasoned that saccades cannot be well prepared in a situation where a foveally informative visual stimulus is expected, because attention is probably directed at the center to await the forthcoming S2. For instance, Rizzolatti et al. (1987) suggested that focusing attention to the saccade target precedes the actual execution of the saccade. This preparatory process may be impossible when attention is already directed at the center. However, this process can be performed within the S1-S2 interval when S2 is auditory. Thus, saccade preparation may be easier with an auditory S2. If this is correct, then the use of an auditory S2 increases the possibility of finding lateralized components reflecting saccade-specific preparation. More generally, this manipulation enabled us to examine whether preparation differs as a function of S2 modality. In addition to these modifications, more trials were presented to improve the signal-to-noise ratio and more electrodes were attached at posterior sites, as the parietal lateralized component may be more evident at sites not measured in the study by Wauschkuhn et al. (1997). In order not to miss relevant information, as was possibly the case with a frontocentral lateralization by Wauschkuhn et al. (1997), we decided to analyze the data in the S1-S2 interval by means of a PCA (Donchin and Heffley 1978; Möcks and Verleger 1991), which gives a more objective decomposition of the relevant variance and a more comprehensive overview than the subjective determination of relevant time intervals. Based on the working model, the following predictions can be made with regard to the ERL components during the S1-S2 interval. When S1 indicates the required response direction, then a negative parietal component was expected like the ETPL in the study by Wauschkuhn et al. (1997), as this component was assumed to reflect a supramodal direction code. When S1 indicates the required finger movement, then additionally a negative central lateralization (the LRP), and a later positive parietal component like the LTPL in the study by Wauschkuhn et al. (1997) are expected. When S1 indicates the required saccade, then, in addition to the negative parietal component, a negative frontocentral component may be expected on the basis of the findings of Klostermann et al. (1994). Preparation of a saccade may be more optimal when S2 is auditory than when S2 is visual, which may be reflected in an enhancement of the negative frontocentral component. As a matter of course, no lateralized components are expected when S1 gives no information about the required response or when S1 only indicates the required effector without specifying the response side.
Materials and methods Participants Twelve healthy, right-handed participants performed this experiment. All participants (6 men, 6 women; mean age 26 years) reported normal hearing and normal or corrected-to-normal vision
and had no history of neurological disorders. The Edinburgh Handedness Inventory (Oldfield 1971), indicated that 11 participants were right-handed (laterality quotients +83 to +100), whereas one was ambidextrous (LQ +49).
Stimuli and procedure Visual stimuli were presented on a 14-inch Multisync monitor, and auditory stimuli were delivered by a headphone. Participants were seated in a comfortable armchair in a sound-attenuated, electrically shielded chamber and viewed the monitor from a distance of approximately 100 cm. A choice response task was used with four alternatives. Upon the arrival of S2, participants either had to move their eyes to the left or right saccade target, or to press the left or right key with the index finger of their corresponding hand. The sequence of visual events on each trial for the visual and auditory condition is shown in Fig. 1. These two conditions were performed on two separate days, with their order balanced across participants. Trials started with a white fixation cross (0.75°×0.65°) displayed in the center of the screen for 1500 ms. Next, S1 was presented in the center for a duration of 300 ms. S1 consisted of a white frame (1.4°×0.7°), enclosing a yellow letter (A, H, or X; 0.3°×0.45°) and two red arrow heads (0.4°×0.5°), one at either side of the letter. The letter provided information about the required effector. An A indicated a saccade (A denoting Auge, German for eye), an H (hand) indicated a finger movement, and an X indicated that this information would be given by S2. The arrow heads gave information about the required response direction, by pointing either to the right or to the left. If both arrow heads pointed inward then S2 indicated the required response direction. Thus, S1 provided either full information (both direction and effector; e.g.,>H>), or direction information (e.g.,
