Figure 6.

Distribution of MEP components between discrete windows of preferred timing. A, Description of the methods used to identify the different subcomponents. EMG signals from one trial of one participant. EMG derivative was thresholded from 5 ms before the onset of the MEP to the end of the positive (up-going) phase of the EMG amplitude trace (black line). The threshold to be exceeded was equivalent to mean baseline EMG activity plus 3 SDs (see Materials and Methods). The identified subcomponents are marked by the triangles. B, Top, Distribution of subcomponent timings with respect to the timing of the subjects' peak MEP amplitude, for both the optimal and nonoptimal phase. *Difference between the two distributions higher than by chance. Bottom, The two distributions are best fitted by a mixture of four Gaussians. One Gaussian was very small and broad. The remaining three Gaussians corresponded to the peaks in the group data and had timings compatible with a series of I waves.