Fig. 4.
Effect of spontaneous activity on response variability in V1. (a) The raw BOLD response is plotted for all trials of a representative participant. Solid line shows the average; shaded area represents the standard error. (b) The spontaneous input into ROIstim as estimated by VOIproxy. Solid line and shaded area, same as in a. (c) Subtracting this estimated spontaneous activity reduced the variability in the evoked BOLD responses. Solid line and shaded area, same as in a. (d) Signal power (top), noise power (middle), and SNR (bottom) for all participants before (left, in blue) and after (right, in red) subtracting the estimated spontaneous activity. Noise power was significantly reduced, and SNR significantly increased, after spontaneous activity subtraction.