FMRI activation related to preparatory set is correlated with saccade latency in human frontal eye fields but not in the supplementary motor area

Abstract

Variation in saccade latency in response to identical sensory stimuli has been attributed to variation in preparatory set. Here we report the first evidence for a relationship between saccade latency and set-related activity in the human frontal eye fields (FEF). Event-related fMRI was used to examine the activation time-courses during a preparatory gap period (2 sessions of 144 trials with 5 subjects), during which no visual stimulus was presented and no saccades were made. The subject simply anticipated the appearance of a flashed peripheral target. Each trial began with the presentation of a central fixation cue (3 s), followed by a green (pro-saccade) or red (anti-saccade) central instructional cue (3 s). This was followed by a 0 (no gap) or 2 s (gap) period of darkness, followed by appearance of a flashed peripheral target (100 ms). Saccade direction and latency were recorded during scanning for each subject. 2-S gap trials were sorted according to short (top 25%) vs. long (bottom 25%) saccade latencies. Examination of the time-courses of activation in the FEF for 2-s gap trials showed a greater build-up in activity during the gap period for short as compared to long latency saccades. In contrast, the supplementary motor area (SMA) exhibited preparatory activity that was not different for short- and long-latency saccades. Replicating our previous work, activation in the intraparietal sulcus (LIP+) did not show preparatory build-up during the gap. These data provide evidence that the FEF contributes to the generation of preparatory set and that such signals may underlie the observed behavioral variability. Moreover, the differences in the pattern of activation we observed in FEF, SMA, and LIP+ demonstrate a functional dissociation between these three primary human oculomotor areas.