Transient Visual Evoked Potentials (tVEPs) to Contrast-Reversing Patterns: A Frequency Domain Analysis

Abstract

Purpose: : To characterize the transient visual evoked potential (tVEP) waveform elicited in humans by contrast-reversal of checkerboard patterns and to identify mechanisms in the frequency domain.

Methods: : In 40 33-s runs per stimulus condition, luminance contrast (100%) of checkerboard patterns was contrast-reversed in time with a 1 Hz square-wave signal. This extensive testing of individuals (N = 10) served to greatly decrease the noise levels in the records. At a binocular viewing distance of 114 cm, field size was 10 deg, and five check widths were used: 4.7-75 minarc in octave steps. The first second of each run was discarded and the remaining 32 s of EEG (synchronized to the display’s frame rate) were divided into four 8-s epochs, yielding a total of 160 epochs per condition. Fourier analysis was performed to obtain the frequency composition of each epoch and a statistical measure, magnitude-squared coherence (MSC), was applied to determine the significance of each frequency component. Principal component analysis (PCA) was used to explore the number and character of underlying neurophysiological mechanisms that contribute to the response.

Results: : MSC analysis revealed that the tVEP is comprised of significant even harmonic components extending to high frequency - in some cases, greater than 80 Hz. PCA yielded six distinct frequency bands that are only weakly or moderately intercorrelated. Repeat sessions on observers demonstrated that the morphology of the time-domain tVEP waveform and the corresponding frequency-domain functions are highly reproducible. Characteristic differences were found in tVEPs among observers, particularly in later time-series deflections corresponding to lower frequency components.

Conclusions: : The presence of several distinct frequency bands in the Fourier decomposition of the transient VEP, which may reflect distinct neurophysiological mechanisms, indicates that this type of analysis might provide selective and sensitive measures of dysfunction in various disease states. The discovery of profound differences in VEP morphology among observers indicates that ensemble-averaging of VEP waveforms from different individuals is problematic.

Keywords: pattern vision • electrophysiology: non-clinical • visual cortex