Dependence of presaccadic cortical potentials on the type of saccadic eye movement

Electroencephalography and cfinical Neurophysiology , 83 (1992) 179-191

179

© 1992 Elsevier Scientific Publishers Ireland, Ltd. 0013-4649/92/$05.00

EEG 91550

D e p e n d e n c e o f p r e s a c c a d i c cortical potentials on the type o f s a c c a d i c eye m o v e m e n t * I. Evdokimidis a, T. Mergner b and C.H. LOcking b " Department of Neurology, Athens National University, Athens (Greece), and h Department of Neurology, University of Freiburg, Freiburg (Germany) (Accepted for publication: 16 March 1992)

Summary Premovement cortical potentials were studied with 4 types of saccadic eye movement: (a) visually triggered saccades of normal reaction time (RT; regular saccades); (b) visually triggered saccades of extremely short RT (express" saccades); (c) saccades towards predicted target locations (anticipatory saccades); (d) saccades back towards predicted location of fixation point (refixation saccades). With all 4 saccade types a "presaccadic negativity" with the maximum at the vertex (Cz) was observed. A bilaterally symmetrical component contained in this potential (being smallest with almost unconsciously performed reflxation saccades and smaller in trained than in naive subjects) appeared to be related mainly to the subjects' volitional effort. In addition, anticipatory and refixation saccades were preceded by an early, widespread contralateral negativity, which we relate to cortical activities that prepare, in general terms, action within or towards the hemifield containing the saccade goal. During the 60 msec before anticipatory saccades, a negativity occurred over the contralateral central lead, which may reflect neural activation in the frontal eye field (FEF) and premotor cortex. In contrast, regular saccades were preceded 30 msec before onset by a negativity over the contralateral parietal cortex, which probably reflects an activation of parietal visuo-motor neurons. No lateralization of the cortical potentials was observed before express saccades, which suggests that these saccades are generated in a reflex-like way mainly by subcortical mechanisms. Key words: Premotor cortical potentials; Saccadic eye movements; (Humans)

Saccadic eye movements are performed in order to foveate the image of a visual object on the retina. There is extensive evidence from clinical and experimental work that the cerebral cortex is involved in the preparation of these eye movements (cf., Wurtz and Goldberg 1989). In humans, several researchers have reported that saccades are preceded by changes in EEG potentials. First, an early developing "premotor negativity" (PMN) is observed, which is followed by a "premotor positivity" (PMP) and, just prior to saccade onset, by a positive "spike potential" (cf., Becker et al. 1972, 1973; Kurtzberg and Vaughan 1973, 1980, 1982; Armington 1977, 1978; Thickbroom and Mastaglia 1985a,b). The "spike potential," which starts about 20 msec prior to, and peaks at, the time of saccade onset, appears to stem mainly from the activity of oculomotor muscles or neurons (Becker et al. 1973; Armington 1977, 1978; Thickbroom and Mastaglia 1985b; Riemslag et al. 1988; Moster and Goldberg 1990).

Correspondence to: Dr. T. Mergner, Neurologische Universit~itsklinik, Hansastr. 9, D-7800 Freiburg (Germany). Tel.: (0)761 270 513. * The study was conducted at the Freiburg University. Supported by Deutsche Forschungsgemeinschaft, SFB 325.

The PMN is reported to start 1 sec or earlier prior to saccade onset and to have a maximum over the vertex (Becker et al. 1972, 1973; Thickbroom and Mastaglia 1985a; Moster and Goldberg 1990). Kurtzberg and Vaughan (1982) considered it to reflect an activation of the frontal eye fields (FEF), whereas Moster and Goldberg (1990), who noticed a slight contralateral preponderance, related it to presaccadic activity in the supplementary eye fields (a subdivision of the supplementary motor area, SMA). In contrast, other authors stressed its resemblance to the contingent negative variation (CNV; prior to stimulus triggered movements) or to the "Bereitschaftspotential" (prior to self-paced movements) and related it to more general preparatory mechanisms for voluntary acts (Becker et al. 1973; Thickbroom and Mastaglia t985a, 1990). The PMP starts about 30-300 msec prior to saccade onset (Becker et al. 1972, 1973; Armington 1977, 1978; Kurtzberg and Vaughan 1980, 1982; Thickbroom and Mastaglia 1985a). Considered an analogue of the PMP prior to voluntary movements of the extremities, which originally has been related to the elaboration of the "motor plan" (Deecke et al. 1968, 1976), the presaccadic PMP has been thought mainly to reflect the activity of parietal visuo-motor centers (Kurtzberg and

180 Vaughan 1980, 1982; Thickbroom and Mastaglia 1985a). However, this potential is bilaterally symmetrical (Thickbroom and Mastaglia 1985a; Moster and Goldberg 1990), a fact which is difficult to reconcile with the idea of a specific presaccadic function. Furthermore, distinction of the PMP from resolving negativity of the preceding PMN is difficult (Moster and Goldberg 1990). Saccades may be initiated in relation to different behavioral contexts, which could mean that cortical preparatory activities vary in degree and spatial distribution accordingly. Therefore, different saccadic tasks have been applied in earlier studies. The PMN has been found to have larger amplitudes with self-paced as compared to visually triggered saccades (Kurtzberg and Vaughan 1982) and with visual stimuli having a predictable (3 sec stimulus intervals) as compared to an "unpredictable" (2.5-3.5 sec intervals) timing (Thickbroom and Mastaglia 1985a). The PMP, in contrast, shows larger amplitudes with visually triggered, as compared to self-paced, saccades (Kurtzberg and Vaughan 1982), and it has steeper slopes with predictable as compared to unpredictable timing of the visual stimulus (Thickbroom and Mastaglia 1985a). Noticeably, no considerable topographic differences have so far been demonstrated with different saccadic tasks, although such differences might be expected from neuron recording in animals (cf., Goldberg and Segraves 1989; Andersen and Gnadt 1989). In fact, it appears surprising that there should be no clear potential changes contralateral to the saccade goal over the frontal and parietal eye fields, which, according to animal work and clinical experience, are the most important cortical centers for saccade preparation (cf., Wurtz and Goldberg 1989). We therefore reinvestigated the presaccadic cortical potentials in a CNV-like paradigm, comparing between different types of saccadic eye movement and searching for hemispheric preponderances ("lateralizations") of the potential changes in relation to the saccade type. In particular, we distinguished on grounds of saccadic reaction time (SRT) and stimulus condition between (i) regular saccades (visually triggered, normal SRTs), (ii) express saccades (visually triggered, extremely short SRTs; cf., Fischer 1987), (iii) anticipatory saccades (self-initiated, towards predetermined target location), and (iv) refixation saccades (performed, often unconsciously, after the task, towards the location where the fixation point was expected to reappear). Express saccades are particular in that they occur in normal subjects mainly in conditions where fixation is disrupted just prior to target appearance (Fischer 1987) or attention is withdrawn from the fixation point (Mayfrank et al. 1986). An abnormally high percentage of express saccades is found in patients with frontal lobe lesions (Guitton et al. 1985; Braun et al. 1989). These reflex-like saccades appear to be mainly of sub-

I. EVDOKIMIDISET AL. cortical origin; in monkey their occurrence depends on the integrity of the superior colliculus (Schiller et al. 1987). In contrast, anticipatory saccades (which precede target onset and usually occur if the subject knows location and time of target appearance) and regular saccades involve cortical preparatory mechanisms. The experiments were performed in two different groups of subjects, one being "naive" and the other "trained" with respect to these kinds of visuo-oculomotor experiment. We hoped that an inter-group comparison would help us to distinguish between presumed unspecific effects related to the subjects' volitional effort (supposed to be higher in naive than in trained subjects) and effects more specifically related to the cortical preparation of saccades (presumed to be similar in both groups).

Methods

Subjects Sixteen subjects, ranging in age from 24 to 49, with normal vision and eye movements participated in the study. They represented 2 groups of 8 subjects each. Those in the first group (3 female and 5 male students) had never before participated in similar studies ("naive subjects"; 7 right and 1 left handed). Those in the second group (2 female and 6 male members of the clinic staff and students) had participated earlier in eye movement studies ("trained subjects"; 7 right and 1 left handed).

Stimulation and procedure Subjects were seated in a dimly illuminated room 72 cm in front of a computer display on which the visual stimuli were presented. The display showed in the center of a dark green background (15 c d / m z) a white spot (0.2° of visual angle; 60 c d / m 2) which served as fixation point. A red spot (0.3°; 100 c d / m 2) served as target and was presented randomly at an eccentricity of 8° either on the right or left side. Subjects were instructed to view attentively the fixation point and to perform a saccade towards the target light immediately following its appearance. The temporal characteristics of fixation point and target presentation are depicted in Fig. 1A. In this figure, tl and t2 give the durations of fixation point presentation and extinction, respectively, while t3 gives the interval between the onsets of fixation point and target presentations, and t4 the duration of target presentation. Since tl was varied randomly by 200 msec and was shorter than t3 by at least 100 msec, there resulted a temporal gap of 100-300 msec between the extinction of the fixation point and the appearance of the target ("gap paradigm"). Total trial

PRESACCADIC POTENTIALS AND SACCADE TYPE A FIXATION POINT

:

t~ =1950-2150 ms

:, l=9

181

t2 =3000 ms

I AP= ,oo.3ooms)

I

TARGET :4 t3=2250ms

:

!

t4=1500ms

', ',

B HOR.EYE POSITION

ri

I

Ant. Sacc.~;'i,~(---Reg, Sacc. L ~ Exp, Sacc.

Je

In both the " s t a n d a r d " and the "predictable" conditions we also evaluated the potentials prior to saccades which the subjects performed after the task back towards the location where the fixation point would r e a p p e a r about 3 sec later ("refixation saccades"). Unlike the aforementioned saccades, these refixation saccades were centripetally oriented. In addition, centrifugal refixation saccades were obtained in a "predictable" condition where the fixation point was displayed at an eccentricity of 8 ° and the target appeared at the center

("eccentric" condition). Cz

5 .~V

Fig. 1. Schematic presentation of the experimental paradigm. A: temporal sequence of stimulus presentation, tl and t2, time periods of fixation point appearance and extinction, respectively: t3, time interval between fixation point and target appearance; t4, time period of target presentation. Note that variation of tl leads to a temporal gap of 100-300 msec between the extinction of the fixation point and the target appearance. B: schematic depiction of the different saccade types that were evaluated (Reg. Sacc., regular saccade; Exp. Sacc., express saccade; Ant. Sacc., anticipatory saccade; Ref. Sacc., refixation saccade. C: overview of the cortical potential changes obtained with these stimuli (the presaccadic spike potential is obscured since this average was stimulus triggered).

duration (sum of tl and t2) varied between 4950 and 5150 msec. This gap paradigm with randomized target location represented our "standard" stimulus condition. The gap, which disrupts visual fixation prior to target appearance, is known to yield a bimodal distribution of the saccadic reaction times (SRTs), the first peak at about 110 msec representing "express saccades" and the second peak at about 150 msec representing "regular saccades" (cf., Fischer and R a m s p e r g e r 1984; Fischer 1987). In addition, a few anticipatory saccades (SRT < 85 msec) were found with the " s t a n d a r d " condition (compare Kalesnykas and Hallett 1987), but they were not considered for analysis. Rather, anticipatory saccades were taken from a modified version of the " s t a n d a r d " condition. Target location, instead of being randomized, was kept constant for 30 successive trials. Note that, in this condition, both the location of the target is known beforehand and the time of its appearance is announced by the extinction of the fixation point ("predictable" condition). As a consequence, the S R T distribution was essentially monomodal with a peak at about 0 msec ( = target appearance). The saccades within the S R T range of 300 msec before to 85 msec after target appearance were considered anticipatory, while saccades outside of this S R T range were not taken for analysis.

To allow comparison of a purely visual task with the above visuo-oculomotor tasks, a 4th condition was used; it represented a repetition of the " s t a n d a r d " condition except that the subjects were instructed to view attentively the display without performing any eye movement ("no mouement" condition).

Recording The E E G was recorded with conventional techniques using a time constant of 10 sec and a low-pass filter of 75 Hz. Records were from Fz, Cz, Pz, Oz, F3, C3, P3, O1', F4, C4, P4, and 0 2 ' versus linked ear electrodes. For the sake of an equidistant inter-electrode spacing, we used a more lateral (more occipitotemporal) placement of the occipital electrodes than usual (O1' and 0 2 ' ) . The amplified E E G was written on a polygraph for on-line control. Furthermore, the E E G was digitized at 500 Hz, displayed on-line on a computer monitor, and stored simultaneously on hard disk for off-line analysis. Recording and storage comprised, in addition to the E E G , the horizontal and vertical eye positions (measured with the infrared Iris ® system, Skalar), an analog code of fixation point and target presentation, and an analog signal of breathing measured from the respired air by means of a thermistor. During the initial experiments, we noticed that our subjects tended to synchronize their breathing with their task performance. The tendency was overcome by prolonging t2 every fifth to ninth trial by 1-3 sec, thus preventing contamination of the presaccadic potentials by scalp potentials related to breathing (cf., Gr6zinger et al. 1974). Records were made in blocks of 3 - 4 min periods, followed by 1-5 min breaks, over a total time of 45-60 min. Varying the sequence of the stimulus condition, 2 - 4 recording blocks were obtained for each condition at different times during the recording session.

Data analysis During off-line analysis, the recording epochs of each trial were displayed on the computer monitor. We could then control for artefacts (eye blinks etc.) and evaluate the S R T (classification of saccade type) and saccade direction. Trials with saccades in the wrong

182

I. EVDOKIMIDIS ET AL.

direction were excluded from analysis, as well as trials with SRTs > 350 msec. The selected epochs were stored in different buffers separately for the two different saccade directions and the different saccade types. If not specified otherwise, the averages were triggered on saccade onset, and the epoch of analysis was restricted to 1200-20 msec before saccade onset (i.e., up to the appearance of the spike potential, which thus was excluded from analysis). The first 50 msec of this epoch was defined as baseline. The number of saccaries per average ranged from 50 to 115 for regular, anticipatory and refixation saccades, while that of express saccades was smaller (19-53). For each of the data sets, grand means were calculated separately for the two subject groups. Three different approaches were used for comparing across the data sets: (1) Since a distinction between PMN and PMP was often difficult or even impossible (compare Moster and

A

Reg. Sacc.

b

a

F

~

~

CO

B

C

Goldberg 1990), we took the integral of the negative potential above baseline (grey shaded areas in Figs. 2 and 5) as a global measure of the "presaccadic negativity." In addition, the amplitude of the potentials was measured for each recording site at 800, 500, 300, 200, 100, 60 and 30 msec before saccade onset. With these data we applied a parametric approach for global analysis of all effects involved (naive vs. trained subjects, saccade type, left vs. right saccade direction, ipsilateral vs. contralateral recording sites, anterior vs. posterior recording sites; BMDPSV TM, Statistical Software Inc., analysis of variance with repeated measures). Significant differences were verified using non-parametric tests (e.g., Kruskal-Wallis test, SAS procedure NPAR1WAY). (2) Since statistically no differences of the presaccadic potentials were observed for right vs. left saccades (cf., below), the data were pooled so as to obtain ipsilaterai, midline, and contralateral records. Then,

~

Z

IP

Exp. Sacc.

CO

Z

IP

CO

Z

IP

b

Ant. Sacc. b

Fa

CO

Z

IP

CO

Z

IP

10 p.V

1200ms Fig. 2. Presaccadic potentials obtained with 3 different saccade types. A: regular saccades. B: express saccades. C: anticipatory saccades. Right and left saceades were averaged such that the left column in each panel gives contralateral (co) and the right column ipsilatateral (ip) recording sites (with respect to saccade goal; z, midline sites). Negativity with respect to baseline is highlighted by grey shading (a, naive and b, trained subjects).

PRESACCADIC POTENTIALS AND SACCADE TYPE

183

"lateralization" curves were calculated by subtracting the ipsilateral from contralateral potentials (the resulting curves c o r r e s p o n d to bipolar records from contralateral vs. ipsilateral leads). (3) Conceivably, these "lateralization" curves do not allow differentiation between a contralateral negativity

A

Reg. Sacc. a

and an ipsilateral positivity, for instance. W e therefore displayed the a f o r e m e n t i o n e d samples at 800, 500, 300, 200, 100, 60 and 30 msec in addition as isopotential maps (Wingz T M for Macintosh, Informix Software). This was d o n e by converting the potentials to a 15-level gray scale code. Deviating from similar approaches in

b

CO:

Fco-Fip

,p: CO:+

Cco-Cip~-"+~-+++++~~,~, .:.4 +~. . . . . . ~'-'" P C O - Pip ; .-~.

~-I

,~

, - F ' + ,

•

• F'

,

Oco-Oip~~~~~ r a Exp. Sacc. a ~,ll]LlJ I ~.LI,,. i,

,

I

J I

r"-'rT~r'

. ~

i

I]-7

L-t

I

,T'-I

I,.d

1 I I

r.-iT,-,7

II1

i

.I/~

r'-,-

2 uV

4 0 0 ms

Fco-Fip

i

b

, I! ~_

,, rTl-i i I l--lq-i ii-111qT~

~ ~ m4z.z_qZ~Ll.kl XI_ ,',~,J~lll~lllliFl[~f

C c o - C i p ~ Pco-Pip ~

-

i m

• t,4.1J~[ L, l_ Oco-O~p~ 'TM i ' l ' r [ @ @

b

C Ant. S,acc.,a ~

Fco-Fip ~ 4 " ] ~

I' i, ',',',"

I

0co-0ipI-HHJ-,~ 4 - ~

'

Cco-Cip!-~ Pco-Pip

ql,l,,llJl,,,tl i,,i,

Fig. 3. Lateralization curves of presaccadic potentials obtained WiThtegt~rar saccades (A), express saccades (B), and anticipatory saccades (C). Upward deflection indicates preponderance of contralateral (co) negativity or of ipsilateral (ip) positivity, downward deflection indicates the opposite. Thin vertical bars, + S.D. Thick horizontal bars, instant of target onset (dispersed, due to variations of SRT); a, naive and b, trained subjects).

184

I. EVDOKIMIDIS ET AL.

800

300

A

~':iii~i

ragr"e- ~ s , .

100

~

~

::::::::::::::::::::::::::::::i:i:!:i:i:i:!:i:i:i:i:i:~!! -"...... ~--.~ ,'+.-~.--..'-'-':..:

:' : ~ . .

,'-:':.':T:T..','.'i':T:'~':'i . ' ; " . ' • = ' " : ' " " ="';••'='-=•":

.

.

.

.

.

.

"

~

,",',",',',"

""-'"'='"=

"

.

. . . .

...'-i': "'"'-"";"

='"

.

.

I,,.., X

e'..~__

b a

,..,, I,,' CC

.

.

'

=

:

~

'

"

"="'= '" " " " " " ' " " ' " ; : ' ; "

i!#ii!i!!ii i "- ~"~::;;~,"";' "~.,'2'*':""~""~":~:

I:::

30 ms

•: . ~ , ' , ' , ~ , ~ , ' , " ~,:-~-:.i-:

':==":':"

~='....~'..?..=i=i~i~il ~ i / I

60

~ {~}~N~:.":i.~::~.":#~~:i~N ~,.'~:~.'.':

.~....~...k~',/, ",/,," ~,..'...~

[ : ~ : : ~ !

............................... ,,, --..... :::::::::::::::::::::::::::::::::::::::::::::::::::

=:~'~.~.='.-',-,~~,~.~! ..;..,~ "

~m'.~.,~'

,

,

............................................

I

,

"s " ~ .......................................... ~.,, :::::::::::::::::::::::::::::

. . . . .

m':.,~

=~

,

..........

,

-

,

~

4

[-...~..~.~,," ~" ,, ;,,;,,~,,; ,,' ....,

,,-:.,,,,, .

.

.

= : :

";.;::':':

,. •

"," .......

.': :::::::::::::::::::::::::::::::::::

.......

.,. ,,, ,,,, % *,,, ,~ ,,,

=" ........ • .,"

.

.

.

.

.

.

.

.," ,.' ,,' .

:.,,,..'..,.-.:.:,-.;,-,:.-::--..,:.-,:.-v

., . . . .

.,-.,

=-..

".'..,. ",.'°'..'..,', .,.- ,..-.-...,'.....

~

~.~~

C i::::::::::::::::::::::::::::::::::::::::::: -

i!)?i?i?)i;?i?)?)?i?)?)?;?iii?)?ii)iiiii??i?;i):il)~):??:~?i)???)?)??ii?)~ i!~i!51ilii5~[iiii%!i~!!i!iiiiiii12111ii)i152!272115[ii:5151!~ii!!!ii ::::::::.: ::::::::::: : : : : :::::5:::::::: : :::::::::::::::::::::

iiii::ii~i~ii::!

:::::::::::::::::::::::::::::: : : ::5:::::::::::::

:::::::::::::::::::::::::::::::::::::::::::::::::::::::: :::::::::::::::::::::::::::::::::::::::::::::::::::::::: :.:;::~;';- ~ ';;::::::: ::::::::::;5:

::i~::~:i~:=i~:i~21si;i~,iii!iiiii!iiiiiiiiiii!iiii

i)i?iiiii)iii)~)~)ii~ii)i)iii}i;i)i)~}~)~?i)~)ili)iii)i)ii~)~)ilili)i)i;il

D Ant. Sacc.

".:,i':'~': :-'-:'~i~.~ ... ~.-"'~' :!--" .4 ...... .~.':;.~.'.~.i

::::::::::::::::::::::::::::::::::::::::::::::::::::::: ",::-'.,'.. :::5- e" ,," " ;

~. ".':, ~:"~:":.'; ":;;:&'.~.i:'..7.

•~.i.~.i.~-i:~i::#:ii~,:;:il ."i:i ",',',',';!':T:'i':T:'i'

:::::::::::::::::::::::::::::::: : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : ,.,

",',',',';i':'i':T:T:T

•

",,',,'.,',,';!':'i':T:'~':'~" , ;.."..-"..."..." , , , I . , , ; , , , l , , , ; ,•.T:T:T:T:T ,;.,,;.*.i,,,;-

i-;

• ;,

... % ... % ... ; • . . : . . . : - . : . . /........e .e :.:.:.:.

:::::::::::::::::::::::::::::::

~ ,

~

~

•

... =. ; . . . ; . . . : . . ~ . . - . : - . . : ~ . , .

':'";'YT;::"."::'".'.'"'"

.1.;-;o;..*.:.:.:.:

::,;'*,.

,.,, -•. -,,. ,,. ,,,. ,,,..,.

,

~ ' . • l = ";,

% .1-,..,.;-.

~ ,:,, .,;..-~..~

• .~

-5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0 0.5 1.0 1.5 2.0

2.5

E

Fig. 4. Presaccadic potentials depicted as isopotential maps. Figures on top indicate time (msec) prior to saccade onset. Grey scale range in microvolts. A: regular saccades, and B: express saccades (naive subjects). C: potential differences across regular and express saccades (obtained by subtracting the latter from the former; pooled data of naive and trained subjects). D: anticipatory saccades (naive subjects). E; potential differences across anticipatory and express saccades (pooled data).

PRESACCADIC POTENTIALS AND SACCADE TYPE

earlier studies (e.g., Thickbroom et al. 1984), the recording area (F3-F4 to O 1 ' - O 2 ' ) w a s treated as a flat rectangle (cf., Results). Furthermore, differences across saccade types were visualized by subtracting the corresponding isopotential maps from each other.

185

tively). This applied only to naive subjects, whereas no difference was found for trained subjects. Also, there was no difference between regular and express saccades in either of the two subject groups. The following presentation will be restricted to the lateralization curves and isopotential maps, rather than giving detailed descriptions of the potential wave forms.

Results

Regular saccades The mean percentages of regular, express and anticipatory saccades in the "standard" condition were 85, 13 and 3% and 57, 37 and 6% in the naive and the trained subjects, respectively. In the "predictable" condition the corresponding figures for anticipatory saccades were 57 and 58%. With the illuminations used in our experiments for background, fixation point and target, the SRT range of express saccades was 85-144 msec and that of the regular saccades was 145-350 msec. A slowly rising negativity, starting with the onset of the recording epoch, was observed with regular, express and anticipatory saccades in both subject groups. It represented a CNV-like potential (i.e., the appearance of the fixation point, S1, started the task and announced the appearance of the target, $2; cf., Rockstroh et al. 1982). An example of this CNV-like potential is shown in Fig. 1C (recording from Cz; regular saccades). Note that the appearance of the fixation point is followed after a visual response by a slowly rising negativity (the presaccadic negativity), and that saccade execution is followed after a visuo-motor response ("lambda wave"; cf., Barlow and Cigfinek 1969) by a slow decay towards baseline. Not shown in the figure are the potentials that preceded the refixation saccades; they were different and will be considered separately in a later section. When comparing the presaccadic negativity (its integral; cf., Methods) across leftward and rightward saccades, we found no statistical differences for subject groups, the different saccade types and the recording sites. We therefore pooled the data of right and left saccades. We distinguish in the following only between contralateral, midline and ipsilateral recording sites. Fig. 2 shows the resulting grand average curves for regular (A), express (B), and anticipatory (C) saccades for naive (a) and trained (b) subjects. The presaccadic negativity (global value for all recording sites) was found to be larger in the naive than the trained subjects by a factor of 1.2, 1.9 and 2.1 for regular, express and anticipatory saccades, respectively ( P < 0.01; P < 0.005; P < 0.005). In a corresponding inter-condition comparison, the presaccadic negativity before anticipatory saccades ("predictable" condition) was higher by a factor of 1.5 and 1.4 than before regular and express saccades ("standard" condition; P < 0.005 and P < 0.01, respec-

No lateralization of the presaccadic potentials was found for this saccade type in both naive and trained subjects up to 100 msec before saccade onset (Fig. 3Aa,b). During the last 60 msec, the lateralization curves showed a rise, indicating that contralateral leads were more negative than ipsilateral ones or ipsilateral leads more positive than contralateral ones. The effect was significant ( P < 0.005) in both naive and trained subjects, and was more pronounced over occipital and parietal as compared to central and frontal sites. In the corresponding isopotential maps of the naive subjects (Fig. 4A), the lack of an early lateralization was reflected in a bilaterally symmetrical negativity with the maximum at Cz in the 800, 300 and 100 msec templates (the same applied to the 500 and 200 msec templates; not shown). At 60 msec, the negativity became attenuated, most prominently over ipsilateral occipital and frontal leads. These changes continued at 30 msec, with some negativity remaining over the midline central and the contralateral parietal leads. Qualitatively similar isopotential fields were obtained with the trained subjects.

Express saccades No lateralization was seen with the express saccades both in naive and in trained subjects (Fig. 3B; the relatively large S.D. values are essentially due to the rather small number of express saccades of the individual averages). Fig. 4B shows the corresponding isopotential maps of the naive subjects; there is a bilaterally symmetrical negativity with the maximum at Cz, which attenuates in the 100-30 msec samples, most prominently over posterior sites.

Anticipatory saccades An early starting lateralization was observed with these saccades both in the naive and trained subjects (Fig. 3C). It consisted of a rise which was more pronounced over the central and frontal as compared to the parietal and occipital leads. This difference was significant ( P < 0.005) with both the naive and the trained subjects. In the isopotential maps of the naive subjects (Fig. 4D), the lateralization was reflected by a pronounced contralateral negativity with the maximum at the central lead. This negativity declined somewhat during the last 100 msec prior to saccade onset, with the ipsilat-

186

I. EVDOKIMIDIS ET AL.

A

a

b

F C P 0 CO

Z

IP

Ba

b

~ ~

P~-~ 0

~

CO

~

~

,~~~ Z

IP

CO

Ca

C

\

~ - ' Z- - / ' ~

~ IP ~

z

Ip

b

~

CO

Z

co

IP

jp.v

1200 ms Fig. 5. Presaccadic potentials obtained with the refixation saccades (a, naive and b, trained subjects). A: "standard" condition. B: "predictable" condition. C; "eccentric" condition. Presentation as in Fig. 2.

Inter-saccade comparisons

eral frontal a n d occipital l e a d s even b e c o m i n g positive. A t 30 msec, t h e r e was r e i n f o r c e m e n t o f the negativity o v e r t h e c o n t r a l a t e r a l c e n t r a l lead. E s s e n t i a l l y similar m a p s w e r e s e e n with the t r a i n e d subjects.

U s i n g the 800, 500, 300, 200, 100, 60 a n d 30 msec s a m p l e s for i n t e r - s a c c a d e c o m p a r i s o n s (cf., M e t h o d s ) , t h e r e was only o n e d i f f e r e n c e b e t w e e n r e g u l a r a n d

a

b~

l

CO:Jp +

Fco-Fip

l i p :co:+

Cco-Cip

Pco-Pip Oco-Oip

-~.----; ~~=;-'--~'~--

"

~

pV .

.

.

.

.

.

.

.

.

.

.

.

.

.

600 ms

Fig. 6. Lateralization curves obtained for the refixation saccades (a, naive subjects, b, trained subjects). Full lines, "standard" condition; dashed lines, "predictable" condition; dotted lines, "eccentric" condition. Presentation as in Fig. 3.

PRESACCADIC

POTENTIALS

AND

SACCADE

TYPE

187

express saccades which was statistically significant (P < 0.005) in both naive and in trained subjects. It was the negativity over the contralateral parietal lead in the 30 msec sample seen with regular and not with express saccades (compare Fig. 4A and 4B). This difference is reflected in the isopotential maps of Fig. 5C, which were obtained by subtracting the latter from the former samples for the pooled data of naive and trained subjects•

800

500

Anticipatory saccades differed significantly from regular and express saccades in both trained and naive subjects with respect to the lateralization (contralateral preponderance of negativity) in the 300-30 msec samples (cf., Fig. 4D). This effect is illustrated in the isopotential maps that were obtained by subtracting the express from the anticipatory data. In the resulting maps (Fig. 4E; pooled data of naive and trained subjects) there is an early widespread negativity over the

1 O0

300

30 ms 2.5 -2.0 .1.5 •1 •0 .0.5 0 0.5 1.0 1.5 2.0 2.5 3.0

A

3.5 4.0 4.5 5.0

B

,.,, . . .

-5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0

•

•,w, ,,,,_

• •

:i~ii!#iiiii:ili:iiiiii}iiii:ili!i!:i!i~ii~ii~ii~iiiii~i!i!ii!~i~i!i =~::::w:iiiiiii!i!i!!!i~!~i!Zii~ii~iiiiii~iiiiii!!ii!!i!!!~i!!~i

ol.5

........q ii!iiiiiiiiiiiiiililiiiiiiiiiiiii]

!~iiii~iiiii,:ii!? I

-1.0 -0.5 0 0.5 1.0 1.5 2.0 2.5

~ •

......!~ii

C

~iiiiiiiiiiiiiii@ii@iiii ~/iiiiiiiiiiiiiiii~iiiiiilil..,-.i~iiiiiiiiiiiiiiil ~::::!!~,:'..~:~::~l::;,..i.~i~i,~:~:~:~:i ~.~:~.::::.¢:.¢.:~.~.~ ';.~:::.~:: ::: u:..:.'k:......

2"-..•'-... ~. ~ ~,.. :. :. ! •':.;

'..':."...2.i.: , / ' , ~ ~.: ~:.!::. ;(:.~.:.~."%e~,~z~¢ •

•

...••..•°

°

•

%

,..

================================ =========================================

%

•

% .:::::::::::

. :::::;:::

.%

%

,~..:.., -....:::~.~ %~.:.:.:..:::..

:.i.:.i.:.i.:.i.z,q":.q..:i.:. :::::::::::::::::::::

......-, - . ! . • . ~. ~. - , -..,

ili;iiiiiiiii!!iiiiiiiiiiiii i iiiiiiiii!iiiiiii

......>.~$, .., ... ,., ..,:;.~.-., :.:.;,., ......'......

:..'F-.F..: • x ,..F.-F..F. ...LLL:.L:;:."••.iLL!.~.LL L~ :,'::'-'F"!'..F " F " " . . F . : : . : " .

~:::w:: . . . .

iiiliY .

,

iiL.

•

i~!......

•~ iiiiiiiii? : ~" ~x:::::::::b

." ".~ ~

•

+:::::::::::::

============================= s'

~:Z,!:), • . . ...... , ,• .'::::::::::::::::: : .,. ,.. ,a

?iill z :!i!!!iiii!:i•~,•

.

%

:5::::::

.." .,......

.."

,,"

::$~:::::::::.,.:. ::::::::::::::::::::::::::::::::::::::::::::::::: ;'. " - ~ • ~ • i - "• ~ ". -'. ~.-'._".

D

IIIIZZ:II

i i

i~

•

ii~/i#}i::::iiiiiiiiiiiiiiil::iiiii::~ii~::.i::::... Fig. 7. Isopotential fields obtained with refixation saccades (pooled data from naive and trained subjects). A: "standard" condition. B : "eccentric" condition. C: difference between A and B. D : "positive shift," resulting from subtraction of the anticipatory saccade data from the refixation saccade data in the "eccentric" condition• Presentation as in Fig. 4.

188

|. EVDOKIMIDIS ET AL.

contralateral leads in the 300-60 msec templates, and, in addition, a pronounced negativity over the contralateral central lead in the 30 msec template.

Refixation saccades The potentials before the refixation saccades clearly differed from those before the other 3 types of saccade. A positive going shift was seen with all 3 stimulus conditions, being more prominent over anterior than posterior leads and over ipsilateral than contralateral leads (Fig. 5); 300-600 msec after onset, the shift became less steep or even changed direction, as if negative going shifts superimposed on the positive ones. This was most pronounced in the "eccentric" condition (Fig. 5C), which was particular in that the refixation saccades were centrifugal (cf., Methods) and all subjects r e m e m b e r e d to have performed these saccades, as revealed by retrospective questioning. The superposition was least pronounced in the "standard" condition (Fig. 5A) in which the refixation saccades were centripetal and none of the subjects r e m e m b e r e d these saccades. The "predictable" condition was intermediate in these respects (Fig. 5B; centripetal refixation saccades, r e m e m b e r e d by 10/16 subjects). In view of the complexity of these potentials no attempt was made to compare the presaccadic negativity, or some positive analogue of it, across subject groups or stimulus conditions. Rather, we restricted the analysis to comparisons of lateralization curves and isopotential maps across the 3 stimulus conditions. The lateralization curves were found to be similar in the 3 conditions, consisting of an early upward deflection

A

Fco

L

Cco

--

--

--

I

,, I

Stimulus-locked potentials The above-described contralateral parietal negativity before regular saccades might represents a visuo-

B

C

Fco-ip=---

AF

TARGET SACCADE

L ,

~

-

Cco-ip

I

I

with a clear preponderance over anterior versus posterior leads (Fig. 6). They were remarkably similar to those seen with the anticipatory saccades (compare Fig. 3C). Fig. 7 shows the isopotential maps obtained with the refixation saccades for the "standard" and the "eccentric" conditions (A and B, respectively; pooled data from naive and trained subjects). In the 800 msec templates of A there is a bilaterally symmetrical positivity with the maximum at Cz, which resembles an inversion of the presaccadic negativity. In the 500 msec to 30 msec templates, the positivity is enhanced and shifted towards ipsilateral anterior sites. In order to reveal differences between the " s t a n d a r d " and the "eccentric" conditions, the maps in B were subtracted from those in A. The resulting maps (Fig. 7C) show a pronounced bilaterally symmetrical negativity with the maximum over Cz, which resembles the presaccadic negativity previously seen with regular and express saccades (compare Fig. 4A and B). For comparison between refixation and anticipatory saccades (both were self-initiated and directed towards a predesignated target location), we subtracted the maps of anticipatory from those of refixation saccades ("eccentric" condition). The result was a widespread, bilaterally symmetrical positivity (Fig. 7D). This finding indicates that the two saccade types were similar with respect to lateralization and presaccadic negativity.

.

.

.

.

-.r

.

i

i

,

',

Pco-ip

I l.v 5

oco-ip-

i

...

I

I -

2.5

-

I

i

I~V

A ~,.,I

~--~_

" "~

--

2 5

'

145

ms

I.v

Fig. 8. Comparison of the stimulus-locked potentials across the "movement" ("standard") and the "no movement" conditions. Trigger for average on target onset, vertical dashed lines; earliest saccade onset, vertical dotted line. A: potentials over contralateral leads of trained subjects (heavy lines, "movement" condition; thin lines, "no movement" condition). B: corresponding lateralization curves (otherwise as in A). C: difference of lateralization curves across "movement" and "no movement" conditions (heavy lines, trained subjects; dotted lines, naive subjects).

PRESACCADIC POTENTIALS AND SACCADE TYPE

motor or a purely visual potential. In an attempt to resolve this question, we compared the stimulus-locked responses between the "standard" condition (in the following section called " m o v e m e n t " condition) and the " n o movement" condition (cf., Methods). To this end the original records were reevaluated, triggering the averages on target rather than on saccade onset. Fig. 8A shows the resultant averages for the contralateral leads from the trained subjects. Following target onset (dashed vertical line) a negative potential occurs with a peak latency of approximately 165 msec (N165; note, however, that the earliest saccade onset is at 145 msec (dotted vertical line), after which the potential is contaminated by eye movement artefacts, so that only the early part of the rising slope can be considered). The potential was found in both the "movement" and the "no movement" conditions over all recording sites. There was, however, a clear preponderance over contralateral as compared to ipsilateral leads, especially posteriorly. This is shown by the upward deflection of the corresponding lateralization curves (Fig. 8B). The deflection was found to be larger in the "movement" than the "no movement" condition; the difference is given in Fig. 8C (subtraction of "no movement" from "movement" curves). Statistically, the difference was highly significant for the parietal and occipital leads ( P < 0.005, respectively). Qualitatively similar observations were made in the naive subjects (dotted curves in Fig. 8C). This difference could by no means be ascribed to differences in absolute potential amplitude. Note from Fig. 8A, for instance, that the presaccadic part of the N165 was similar or even slightly larger in the "no movement" than the "movement" condition. Also, the N165 at midline leads (essentially free of eye movement artefacts, because of pooling left and right saccade data) were similar in the two conditions (not shown).

Discussion In search for specific presaccadic cortical activity in the human E E G we looked for topological differences across different saccade types, assuming that these would involve the cortical eye fields to different degrees. Also, we conceived that these activities might be masked by a large CNV-like potential mainly related to the subjects' volitional effort. Effort-related potentials were expected to be smaller in trained as compared to naive subjects, whereas saccade-related potentials were expected to be similar in the two subject groups. Clear topological differences were indeed found for different saccade types. In both naive and trained subjects regular saccades were preceded by a negative

189

potential over the contralateral parietal lead at 30-60 msec before saccade onset. Up to then the presaccadic potential was bilaterally symmetrical. No such parietal potential was seen with express and anticipatory saccades. We assume that it reflects an activation of visuo-motor centers in the posterior parietal cortex that deal with the selection of the saccade target. Neurons in this region discharge in a direction-selective way prior to purposeful, visually guided saccades, their activity being in part motor-related (Mountcastle et al. 1975) and in part sensory-related (Robinson et al. 1978), while certain subgroups code spatial aspects of the saccade goal (Andersen et al. 1985). Also the parietal potential in our subjects appeared to be, at least in part, visual-related, since it was found on stimulus-locked averaging in both the "movement" ("standard") and the "no movement" conditions (N165). This finding may be analogous to that of Bushnell et al. (1981); neurons in the monkey's area 7 showed an increase in activity in response to an attended visual stimulus, regardless of whether a saccade was performed or not. Furthermore, Fox et al. (1985) found an increase in regional cerebral blood flow (rCBF) in this region in man with visually triggered saccades, which was similar to that previously seen with visual stimulation alone. Yet, in our experiments, we noted a difference between the "movement" and the "no movement" conditions. The contralateral preponderance of the visual response was clearly more pronounced in the "movement" than in the "no movement" condition. We were surprised to find no overt potential over the contralateral F E F region before regular saccades, since neurons in the monkey's F E F are known to fire before visually guided saccades (Bruce and Goldberg 1985). Also, an increase in rCBF has been described in the FEF region of man with visually triggered saccades (Fox et al. 1985). A possible explanation for this discrepancy is that the activity in the F E F before regular saccades is weak (e.g., as compared to that in the parietal field). In fact, visually guided saccades are still found after chronic lesions of the FEF (Schiller et al. 1980, 1987), and there are direct projections from parietal to brain-stem visuo-motor centers (Lynch et al. 1985). Another observation was the absence of any lateralization prior to express saccades. This would be compatible with the notion that this saccade type is generated in a reflex-like way, mainly by subcortical mechanisms (cf., Introduction). Since both express and regular saccades were collected from the same experiments ("standard" condition), one may look for early differences in the presaccadic potentials ( > 60 msec before saccade onset) which would allow a prediction of the saccade type that followed (regular or express). Conceivably, since there was no early lateralization with

190

either saccade type, one has to resort to the bilaterally symmetrical potentials. Only a minor difference was noted; at 100 msec, attenuation of the presaccadic negativity at posterior sites was more pronounced with express than with regular saccades (reflected in a posterior negativity in Fig. 4C, the result of field subtraction). Anticipatory saccades were different from express and regular saccades in that they were preceded by an early as well as a late lateralization. The early lateralization consisted in a widespread negativity over contralateral leads and some ipsilateral frontal positivity. We assume that it is related to cortical mechanisms that prepare action (not necessarily only saccades) in the opposite hemifield (that of saccade goal). The wide spread of this potential suggests that several cortical areas are involved. The late lateralization consisted in a negativity over the contralateral central lead. Possibly, activity in the F E F contributed to this potential. However, in view of cytoarchitectonic and electrical stimulation studies of the FEF in man (cf., Foerster 1936), we would have expected a more anterior location of this activity. Interestingly, however, in the study of Fox et al. (1985) the region of rCBF increase with saccadic eye movements was found to be located immediately anterior to the rolandic fissure, which corresponds approximately to the site of the negativity seen in the present study. Yet one may conceive that, in addition, the premotor cortex was activated in this particular stimulus condition ("predictable condition") and contributed to the potential. Neurons in the monkey's premotor cortex are known to fire with the anticipation of predictable environmental events (Mauritz and Wise 1986; Vaadia et al. 1988). The fulfilment of the saccade task in the "standard" and "predictable" conditions was followed by a further decline of the presaccadic negative potential, during which subjects performed a refixation saccade. With the decline as baseline, the initial part of the presaccadic potential before the refixation saccade represented a positive drift which resembled an inversion of the presaccadic negativity (cf., Fig. 7A, 800 msec template). After 300-600 msec, a superposition by negative potentials became evident. They appeared to consist of at least two components. One was a lateralization which was similar in all 3 conditions tested and closely resembled that previously obtained with anticipatory saccades. The other was a bilaterally symmetrical negativity with its maximum at Cz, the size of which depended on stimulus condition, being smallest in the "standard" and largest in the "eccentric" condition. Since subjects remembered their refixation saccades in the latter, but not in the former condition, one may assume that volitional effort is a major factor which determines the size of this potential.

1. EVDOKIMIDIS ET AL.

Also the size of the presaccadic negativity before regular and anticipatory saccades appeared to be influenced by the amount of volitional effort. This notion, originally proposed by Becker et al. (1973; also Thickbroom and Mastaglia 1990) was supported in our study by 2 observations: (A) The presaccadic negativity was larger with self-initiated (anticipatory) saccades than with visually triggered (regular) saccades. This finding confirms the earlier observations of Kurtzberg and Vaughan (1982) and Thickbroom and Mastaglia (1985a; see Introduction). (B) The presaccadic negativity in both the "standard" and the "predictable" conditions was larger in the naive than in the trained subjects. Finally, in our study we found no left-right difference in amplitude of the presaccadic negativity, as reported by Moster and Goldberg (1990) for some of their recording sites. This discrepancy cannot be explained by different proportions of right and left handed subjects in the two studies (which were similar), but may depend on methodological differences (saccade type, number of subjects, etc.). Taken together, it appears that the seemingly simple motor act of a saccadic eye movement involves a number of complex preparatory mechanisms at different cortical sites, well beyond the classical frontal and parietal eye fields. The limited spatial resolution of the E E G and the temporal and spatial superposition of several cortical activities hamper the identification of these mechanisms in man. Yet we have shown that the human presaccadic potentials contain lateralized components which depend on the saccade type. To date, however, it is still not known where exactly these components are generated and which aspects of saccade preparation they reflect. The authors are grateful to Dr. J. Schulte-M6nting from the Department of Medical Biometry and Statistics, Freiburg University, for performing the statistics.

References Andersen, R.A. and Gnadt, J.W. Posterior parietal cortex. In: R.H. Wurtz and M.E. Goldberg (Eds.), The Neurobiology of Saccadic Eye Movement. Elsevier, Amsterdam, 1989: 315-335. Andersen, R.A., Essick, G.K. and Siegel, R.M. Encoding of spatial location by posterior parietal neurons. Science, 1985, 230: 456458. Armington, J.C. Potentials accompany eye movements. In: S.J. Cool and E.L. Smith, III (Eds.), Frontiers in Visual Sciences. Springer, New York, 1977: 460-470. Armington, J.C. Potentials that precede small saccades. In: J.C. Armington, J. Krauskopf and B.R. Wooten (Eds.), Visual Psychophysics and Physiology. Academic Press, New York, 1978: 363 -372. Barlow, J.S. and Cigfinek, L. Lambda responses in relation to visual evoked responses in man. Electroenceph. clin. Neurophysiol., 1969, 26: 183-192.

PRESACCADIC POTENTIALS AND SACCADE TYPE Becker, W., Hoehne, O., lwase, K, and Kornhuber, H.H. Bereitschaftspotential, pr~imotorische Positivierung und andere Hirnpotentiale bei sakkadischen Augenbewegungen. Vision Res., 1972, 12: 421-436. Becker, W., ttoehne, O., Iwase, K. and Kornhuber, H.H. Cerebral and ocular muscle potentials preceding voluntary eye movements in man. Electroenceph. clin. Neurophysiol., 1973, Suppl. 33: 99-104. Braun, D., Mergner, T. and Weber, H. Patterns of saccadic reaction times in patients with unilateral cortical lesions. Perception, 1989, 18: 545, Bruce, C.J. and Goldberg, M.E. Primate frontal eye fields. I. Single neurons discharging before saccades. J. Neurophysiol., 1985, 53: 603-635. Bushnell, MC., Goldberg, M.E. and Robinson, D.L. Behavioral enhancement of visual responses in monkey cerebral cortex. I. Modulation in posterior parietal cortex related to selective visual attention. J. Neurophysiol., 1981, 46: 755-772. Deecke, L., Scheid, P. and Kornhuber, H.H. Distribution of readiness potential, premotion positivity, and motor potential of the human cerebral cortex preceding voluntary finger movements, Exp. Brain Res., 1968, 7: 158-168. Deecke, L., Gr6zinger, B. and Kornhuber, H.H. Voluntary finger movement in man: cerebral potentials and theory. Biol. Cybernet., 1976, 23: 99-119. Fischer, B. The preparation of visually guided saccades, Rev. Physiol. Biochem. Pharmacol., 1987, 1116: 1-35. Fischer, B. and Ramsperger, E. Human express saccades: extremely short reaction times of goal directed eye movements. Exp. Brain Res., 1984, 57: 191-195. Foerster, O. Symptomalologie der Erkrankungen des Grosshirns. Motorische Felder und Bahnen. In: O. Bumke and O. Foerster (Eds.), Handbuch der Neurologie. Allgemeine Neurologie, Vol. VI. Julius Springer, Berlin, 1936: 0-357. Fox, P.T., Fox, J.M., Raichle, M.E. and Burde, R. The role of cerebral cortex in the generation of voluntary saccades: a positron emission tomographic study. J. Neurophysiol., 1985, 54: 348-369. Goldberg, M.E. and Segraves, M.A. The visual and frontal cortices. in: R.H. Wurtz and M.E. Goldberg (Eds.), The Neurobiology of Saccadic Eye Movements. Elsevier, Amsterdam, 1989: 315-335. Gr6zinger, B., Kornhuber, H.H. and Kriebel, J. Respiration correlated potentials. J. lnterdiscip. Cycle Res., 1974, 5: 287-294. Guitton, D., Buchtel, H.A. and Douglas, R.M. Frontal lobe lesions in man cause difficulties in suppressing reflexive glance and in generating goal-directed saccades. Exp. Brain Res., 1985, 58: 455-472. Kalesnykas, R.P. and Hallett, P.E. The differentiation of visually guided and anticipatory saccades in gap and overlap paradigms. Exp. Brain Res., 1987, 68: 115-121. Kurtzberg, D. and Vaughan, H.G. Electrocortical potentials associated with eye movements. In: V. Zikmund (Ed.), The Oculomotor System and Brain Function. Butterworth, London, 1973: 137-145. Kurtzberg, D. and Vaughan, H.G. Differential topography of human eye movement potentials preceding visually triggered and self-ini-

191 tiated saccades. In: H.H. Kornhuber and L. Deecke (Eds.), Motivation, Motor and Sensory Processes of the Brain. Elsevier, Amsterdam, 1980: 203-208. Kurtzberg, D. and Vaughan, H.G. Topographic analysis of human cortical potentials preceding self-initiated and visually triggered saccades. Brain Res., 1982, 243: 1-9. Lynch, J.C., Graybiel, A.M. and Lobeck, L.J. The differential projection of two cytoarchitectural subregions of the inferior parietal lobule of macaque upon the deep layers of the superior colliculus. J. Comp. Neurol., 1985, 235: 241-254. Mauritz, K.-H. and Wise, S.P. Premotor cortex of the rhesus monkey: neuronal activity in anticipation of predictable environmental events. Exp. Brain Res., 1986, 61: 229-244. Mayfrank, L., Mobashery, M., Kimmig, H. and Fischer, B. The role of fixation and visual attention in the occurrence of express saccades in man. Eur. J. Psychiat. Neurol. Sci., 1986, 235: 269-275. Moster, M.L. and Goldberg, G. Topography of scalp potentials preceding self-initiated saccades. Neurology, 1990, 40: 644-648. Mountcastle, V.B., Lynch, J.C., Georgopoulos, A., Sakata, H. and Acuna, C. Posterior parietal associatkm cortex of the monkey: command functions for operations within the extrapersonal space. J. Neurophysiol., 1975, 38: 871-908. Riemslag, F.C.C., Van der Heijde, G.L., Van Dongen, M.M.M.M. and Ottenhoff, F. On the origin of the presaccadic spike potential. Electroenceph. clin. Neurophysiol., 1988, 70: 281-287. Robinson, D.L., Goldberg, M.E. and Stanton, G.B. Parietal association cortex in the primate: sensory mechanisms and behavioral modulations. J. Neurophysiol., 1978, 41: 910-932. Rockstroh, B., Elbert, T., Birbaumer, N. and Lutzenberger, W. Slow Brain Potentials and Behavior. Urban and Schwarzenberg, Baltimore, MD, 1982. Schiller, P.H., True, S.D. and Conway, J.L. Deficits in eye movements following frontal eye field and superior colliculus ablations. J. Neurophysiol., 1980, 44:1175-1189. Schiller, P.H., Sandell, J.H. and Maunsell, J.H.R. The effect of frontal eye field and superior co[liculus lesions on saccadic latencies in rhesus monkey. J. Neurophysiol., 1987, 57: 1033-1049. Thiekbroom, G.W. and Mastaglia, F.L. Cerebral events preceding self-paced and visually triggered saccades. A study of presaccadic potentials. Electroenceph. clin. Neurophysiol., 1985a, 62: 277-289. Thiekbroom, G.W. and Mastaglia, F.L. Presaccadic "spike" potential: investigation of topography and source. Brain Res., 1985b, 339: 271-280. Thickbroom, G.W. and Mastaglia, F.L. Premotor negativity associated with saccadic eye movement and finger movement: a comparative study. Brain Res., 1990, 506: 223-226. Thickbroom, G.W., Mastaglia, F.L. and Carroll, W.M. Spatio-temporal mapping of evoked cerebral activity. Electroenceph. clin. Neurophysiol., 1984, 59: 425-431. Vaadia, E., Kurata, K. and Wise, S.P, Neuronal activity preceding directional and nondirectional cues in the premotor cortex of rhesus monkeys. Somatosens. Motor Res., 1988, 6: 207-230. Wurtz, R.H. and Goldberg, M.E. (Eds.). The Neurobiology of Saccadic Eye Movements. Elsevier, Amsterdam, 1989.