Abstract
Purpose. We present a method to identify the onset of cortical activity after visual stimulation using multifocal VEP (mfVEP) recordings. The aim of the study was to compare this latency measure with peak latencies of averaged mfVEP traces.
Methods. 30 visually normal subjects participated in the study. Binocular mfVEPs were recorded between two electrodes placed 4 cm above and below the inion. Dartboard patterns with 60 fields were presented within a stimulus field with a diameter of 41 deg. Within the dartboard fields checkerboard patterns with a mean luminance of 86.1 cd/m2 and a contrast of 99.8% were counterphased following m-sequence stimulation. Either mfVEP traces or their squares were averaged across different subsets of the 60 fields (upper and lower hemifield, 5 rings). For the average of the mfVEP traces (mVEP) latency was derived from the peak implicit time around 90ms. For the average of the squared mfVEP traces (sqVEP) latency was defined by the onset of a sudden rise above the noise pedestal.
Results. (1) When averaged across all subjects the sqVEP for the upper and lower hemifields showed a similar onset of cortical activity (48.5ms vs. 48.0ms) while their peak mVEP latencies differed significantly (100.8ms vs. 91.7ms). (2) We found a continuous shortening of sqVEP latencies from the central ring (55.0ms) to the most peripheral ring (46.3ms) while the mVEP latencies did not show any significant differences (around 85ms).
Conclusions. The mVEP latencies for the different parts of the visual field can not be derived from the sqVEP latencies by adding a constant time delay. This discrepancy may be due to the fact that sqVEP latencies reflect pre-cortical processing of retina and optic nerve, while mVEP latencies involve intracortical processing over a timer interval of about 50 ms. The data suggest that sqVEP latency can be used as temporal reference point to distinguish between pre-cortical and intracortical processing in basic research and in the clinical routine.