Fig. 2.

A and D indicate the average ERSPs of the AF3, F1, F3, F5, and F7 channels corresponding to the stimulation site (i.e. the left DLPFC) before (Pre) and after the DLPFC–PAS paradigm (Post0, Post15, and Post30) for the HC group (n = 29) and the TRD group, respectively. The average ERSPs normalized to the baseline frequency power before TMS (from −500 to −100 ms) are presented. The ERSPs at the Pre, Post0, Post15, and Post30 measurement points are shown from top to bottom. B and E indicate ERSP modulation calculated by the average ERSPs after the DLPFC–PAS paradigm (Post0, Post15, and Post30) minus those before the paradigm (Pre) from top to bottom. Color scales for A, B, D, and E represent the value of decibel power (dB). Nonsignificant values are masked in white (the FDR-corrected P-value > 0.05) in ERSP modulation. C and F show the ESs. Color scales for C and F represent Cohen’s d values. The areas enclosed by the dashed line were used for between-group comparisons. G and H indicate a scatter plot of ERSP modulation within the gamma (30–48 Hz) and alpha bands (8–12 Hz) after 100 ms poststimulation, respectively. The filled plots display individual data in the HC and TRD groups. The unfilled circles and error bars represent the mean values and SDs, respectively. There was a significant group difference in ERSP modulation within the gamma from 100 ms to 300 ms poststimulation at Post15; the ERSP gamma modulation was significantly more pronounced in the HC group compared to the TRD group (P-value < 0.05).