Abstract
Abstract: :
Purpose: Human subjects are typically instructed to fixate on a particular location for many visual tasks, for both clinical tests and basic research. Sometimes subjects do not maintain their fixation very well. In multifocal visually evoked potentials (mfVEP), there can be reversals in waveform shape for nearby retinotopic locations due to convolutions in cortex. We hypothesize that the mfVEP is very sensitive to small changes in gaze position. Methods: The mfVEP is recorded from normal subjects with two bipolar occipital electrodes using the Visually Evoked Response Imaging System (VERIS). The stimulus consists of a dartboard pattern of 60 areas, each area containing a checkerboard pattern of 16 checks. The checkerboard in a given area is pattern reversed according to an m-sequence. In order to simulate poor fixation, subjects are instructed to move their fixation point systematically in a clockwise direction between the endpoints of a fixation cross every few seconds (two degree diameter cross = one degree fixation error). Results are compared against a control condition in which they fixate properly. Results: The effects of 0.5 degree fixation error are small, but 1.0 degree fixation error can produce a large decrease in root mean square signal strength (greater than 50%) in the central foveal region (i.e., within 1.3 degree eccentricity). The size of the effect varies greatly between areas within a subject, and between subjects. In the surrounding area (between 1.3 and 2.9 degree eccentricity) there can be dramatic changes in waveform shape, such as an inversion of positive and negative peaks. Beyond 2.9 degree eccentricity the effects are minimal. Conclusion: Small fixation errors can create dramatic changes in mfVEP amplitude and waveform shape within three degrees of eccentricity.
Keywords: 393 electrophysiology: clinical • 406 eye movements • 394 electrophysiology: non-clinical