The mirror antisaccade task: direction-amplitude interaction and spatial accuracy characteristics.

In this study we examined the performance of human subjects in three oculomotor tasks: a visually guided saccade task (VST), a simple antisaccade task (SAT) and a mirror antisaccade task (MAT). The stimulus presentation was identical for all three tasks, and the differentiation of the tasks was based on the instruction given to the subjects. Subjects were instructed to either look at a visually presented target location (visually triggered saccade task), or to look at the opposite direction of the visually presented target (simple antisaccade task) or finally to look at the mirror location opposite to the location of the visually presented target (mirror antisaccade task). The loading of the simple antisaccade task with the addition of the amplitude requirement did not affect the percentage of directional errors but slowed down the onset of antisaccade execution by 19 ms. The amplitude of the directionally correct antisaccade in the mirror antisaccade task showed a significant distortion of the true mirror target location. This dysmetria followed the same qualitative pattern to that observed for the visually guided saccades, i.e., a near-target hypermetria together with a far-target hypometria, but both these features were exaggerated in the mirror antisaccade task. This distorted amplitude modulation of mirror antisaccade amplitude was completely lost in corrected antisaccades that followed a directional error.