Role of latency jittering correction in motion-onset VEP amplitude decay during prolonged visual stimulation

Visual evoked potentials to motion-onset stimulation (M-VEPs) gradually attenuate in amplitude during examination. The observed decline in averaged responses can be caused by decreases in single response magnitudes and/or increased variability in a response delays, that is, latency jittering. To illuminate the origins of the suppression of M-VEPs during stimuli repetition, we used correlation technique to estimate an upper bound of possible latency jittering of single sweeps and we evaluated the effect of its correction on the amplitudes of three M-VEP dominant peaks P1, N2 and P3. During prolonged visual motion stimulation, the variability of corrective latency shifts in the occipital region increased (r = 0.35: 0.44) and the number of single responses corresponding to the average curve declined in occipital and parietal derivations (r = -0.48: -0.62). While the P1 peak amplitude did not exhibit any time-specific behaviour, the N2 amplitude exhibited a significant decay of 29.4% that was partially reduced to 16.6% in the central occipital derivation by the latency jitter and non-correspondence corrections. The strongest attenuation (32.7%) was observed in the P3 amplitude and was less sensitive to the corrections, dropping only to 27.9%. The main part of the response suppression to repeated motion stimulation was caused by amplitude drop and represents non-stationary process that likely correspond to a fatigue model. The rise of variability in latency jitter correction and the reduction in single responses correlated with the M-VEP were significant factors associated with prolonged motion stimulation. The relation of these parameters to a hypothetical veridical response is ambiguous and can be caused by a time shift of the response or by a change of signal-to-noise ratio. Using selective averaging and latency jitter correction, the effect of response suppression was partially removed.