Multifocal topographic visual evoked potential: improving objective detection of local visual field defects.

PURPOSE To investigate the relationships between the pattern stimulation of different parts of the visual field (up to 25 degrees of eccentricity), the electrode position, and the cortical response to improve objective detection of local visual field defects. METHODS The human visual evoked potential (VEP) was assessed using multifocal pseudorandomly alternated pattern stimuli that were cortically scaled in size. Monopolar and bipolar electrode positions were used. The visual field was investigated up to 26 degrees of eccentricity. Twelve normal subjects and seven subjects with visual field defects of different nature were studied. RESULTS Although the monopolar response is heavily biased toward the lower hemifield, bipolar leads overlying the active occipital cortex (straddling the inion) demonstrate good signals from all areas of the visual field tested. The amplitude is almost equal for the averaged upper and lower hemifields, but the polarity is opposite, causing partial cancellation of the full-field VEP. The degree of cancellation depends mainly on latency differences between the vertical hemifields. The bipolar VEP corresponded well with Humphrey visual field defects, and it showed a loss of signal in the scotoma area. CONCLUSIONS The multifocal VEP demonstrates good correspondence with the topography of the visual field. Recording with occipital bipolar electrode placement is superior to standard monopolar recording. To avoid a full-field cancellation effect, a separate evaluation of upper and lower hemifields should be used for the best assessment of retinocortical pathways. This technique represents a significant step toward the possible application of the multifocal VEP to objective detection of local defects in the visual field.