Occipital gamma-oscillations modulated during eye movement tasks: simultaneous eye tracking and electrocorticography recording in epileptic patients

Abstract

We determined the spatio-temporal dynamics of cortical gamma-oscillations modulated during eye movement tasks, using simultaneous eye tracking and intracranial electrocorticography (ECoG) recording. Patients with focal epilepsy were instructed to follow a target moving intermittently and unpredictably from one place to another either in an instantaneous or smooth fashion during extraoperative ECoG recording. Target motion elicited augmentation of gamma-oscillations in the lateral, inferior and polar occipital regions in addition to portions of parietal and frontal regions; subsequent voluntary eye movements elicited gamma-augmentation in the medial occipital region. Such occipital gamma-augmentations could not be explained by contaminations of ocular or myogenic artifacts. The degree of gamma-augmentation was generally larger during saccade compared to pursuit trials, while a portion of the polar occipital region showed pursuit-preferential gamma-augmentations. In addition to the aforementioned eye movement task, patients were asked to read a single word popping up on the screen. Gamma-augmentation was elicited in widespread occipital regions following word presentation, while gamma-augmentation in the anterior portion of the medial occipital region was elicited by an involuntary saccade following word presentation rather than word presentation itself. Gamma-augmentation in the lateral, inferior and polar occipital regions can be explained by increased attention to a moving target, whereas gamma-augmentation in the anterior-medial occipital region may be elicited by images in the peripheral field realigned following saccades. In functional studies comparing brain activation between two tasks, eye movement patterns during tasks may need to be considered as confounding factors.

Figure 1. ECoG signals modulated during saccade trials in patient #3

Time-frequency analysis relative to the onset of target motion demonstrated that significant gamma-augmentation involving electrode #1 in the lateral occipital region at +105 msec, electrode #2 in the superior parietal lobule at +125 msec, electrode #3 in the Rolandic region at +125 msec, electrode #4 in the premotor region at +300 msec and electrode #5 in the medial occipital region at +225 msec. Time-frequency analysis relative to the onset of saccades demonstrated that significant gamma-augmentation involving electrode #1 in the lateral occipital region at −20 msec, electrode #2 in the superior parietal lobule at −35 msec, electrode #3 in the Rolandic region at +5 msec, and electrode #5 in the medial occipital region at +75 msec; thereby, gamma-augmentation at electrode #4 failed to reach significance. Taking into account the response time was 182 msec on average, gamma-augmentations in the lateral occipital region, superior parietal lobule and Rolandic region occurred prior to and during saccades, whereas those in the premotor and medial occipital regions occurred following the onset of saccade.

Figure 2. Transient gamma-augmentation of ocular origin in the anterior temporal region in patient #3

Time-frequency analysis relative to the onset of saccades demonstrated that gamma-augmentation strictly time-locked to the onset of saccades in the anterior temporal region (T1 through T5). Such gamma-augmentation of ocular origin was better noted on common average montage (black arrows), while significant gamma-augmentation of ocular origin failed to be seen on all bipolar derivations except at T4–T5 (black arrowhead). Gamma-augmentations in the occipital (O1 through O5) and parietal (P2) regions were noted on both montages.

Figure 3. Difference in gamma-augmentation between saccade and pursuit trials

The overall results of time-frequency analysis comparing gamma-modulations between saccade and pursuit trials are shown. The locations of subdural electrodes in five patients were superimposed on a brain template. S: electrode sites showing larger gamma-augmentation in saccade trials compared to pursuit ones. P: electrode sites showing larger gamma-augmentation in pursuit compared to saccade trials. S→P: electrode sites showing gamma-augmentation initially larger in saccade trials but subsequently larger in pursuit trials. P→S: electrode site showing gamma-augmentation initially larger in pursuit trials but subsequently larger in saccade trials. Time-frequency matrixes below show that gamma-augmentation was preferentially elicited by saccade trials in lateral and inferior occipital sites, while gamma-augmentation was preferentially elicited by pursuit trials in a polar occipital site.

Figure 4. Gamma-augmentation elicited by initial involuntary saccades following word presentation

The temporal characteristics of gamma-range amplitudes are shown. The x-axis represents time, while the y-axis represents trials sorted according to the response time. Each row shows gamma-range amplitudes at 80–150 Hz averaged across three trials as a function of time. Gamma-augmentation in the polar occipital region was better time-locked to the onset of word presentation, whereas that in the medial occipital region was better time-locked to the onset of initial involuntary saccade following word presentation.