抄録
In an oddball paradigm in auditory MEG and EEG measurements, an averaged response to frequently presented standard auditory stimulus and that to infrequently presented deviant stimulus are compared to study various aspects of automatic processing of auditory information in the brain. This paper discusses methods for presenting the difference between the responses to standard and deviant stimuli obtained by multi-channel recording. Specifically, we compared R(Δ), the RMS (across the MEG or EEG channels) of the difference between the standard and deviant responses and Δ(R), the difference between the RMS of the standard and that of the deviant response. To evaluate the behaviors of R(Δ) and Δ(R). we used a very simple signal model and obtained the averages and variances of these two statistics in two typical situations. The first is where a mismatch response is elicited and in the other, the responses are equal. Based on this analysis, A(R, Δ), the average of R(Δ) and Δ(R) was recommended. The statistical behavior of A(R, Δ) was also discussed.
