A role for the visual cortex in the dynamic interaction of saccades and visual processing is suggested by functional MRI studies, which reveal saccade-related striate (occipital) cortical activity. 1-6 Little is known, however, of the timing relationship between occipital changes and saccades. Occipital changes preceding the eye movement may be consistent with the notion of corollary discharge, 7,8 whereas occipital activity following the eye movement could signify the response of the visual cortex to a new, shifted visual field. In order to study the time course of occipital activity, we have studied electroencephalographic recordings in association with saccades. We concentrated on gamma range EEGs. It is known that gamma range activity represents local circuit binding of neurons engaged in the same task. 9 Our hypothesis was that there is a lawful relationship of gamma range occipital EEGs in relation to saccade timing and direction as the result of reorganization of neuronal binding in the visual cortex in association with saccades. Our hypothesis was based on the explicit suggestion of Morrone, Ross, and Burr 10 that visual space compression occurs during the short duration of a saccade. Previous studies of perisaccadic EEG changes 11,12 applied Fourier analysis, which is not adequate to quantify short-term EEG changes. Additionally, we concentrated on gamma range activity, as frequencies substantially lower could not be used by the brain to convey information of the chronology of neuronal reorganization during a saccade.Ten normal subjects were studied in comparable age range, with the mean age being 38 years. The EOG and infrared eye scan recorded self-paced saccades and eye position. Permanent markers were located 78 cm from the observer, their distance subtending 60 degrees at the eye. Fourteen channels of EEGs were recorded over posterior occipital and parietal channels. Following a test session, the continuous recording EEG and eye movement recordings were visually inspected, and with our software, cursors were placed, bracketing acceptable saccades. Saccades were marked when their velocity exceeded 200 degrees/second, and they were free of ret-

Our objective was to determine perisaccadic gamma range oscillations in the EEG during voluntary saccades in humans. We evaluated occipital perisaccadic gamma activity both in the presence and absence of visual input, when the observer was blindfolded. We quantified gamma power in the time periods before, during, and after horizontal saccades. The corresponding EEG was evaluated for individual saccades and the wavelet transformed EEG averaged for each time window, without averaging the EEG first. We found that, in both dark and light, parietal and occipital gamma power increased during the saccade and peaked prior to reaching new fixation. We show that this is not the result of muscle activity and not the result of visual input during saccades. Saccade direction affects the laterality of gamma power over posterior electrodes. Gamma power recorded over the posterior scalp increases during a saccade. The phasic modulation of gamma by saccades in darkness--when occipital activity is decoupled from visual input--provides electrophysiological evidence that voluntary saccades affect ongoing EEG. We suggest that saccade-phasic gamma modulation may contribute to short-term plasticity required to realign the visual space to the intended fixation point of a saccade and provides a mechanism for neuronal assembly formation prior to achieving the intended saccadic goal. The wavelet-transformed perisaccadic EEG could provide an electrophysiological tool applicable in humans for the purpose of fine analysis and potential separation of stages of 'planning' and 'action'.

No abstract