Fig. 1: Data Collection and Processing Pipeline.

A Data were collected from 16 MDD subjects using a 64-channel TMS-compatible EEG headcap during a single stimulation session. The protocol consists of rhythmic stimulation trains at random frequencies, followed by an approximate five-minute rest period, after which trains of patterned stimulation are administered. Rhythmic stimulation trains consisted of 40 pulses, with frequencies between 3 and 18 Hz with 0.2 Hz steps administered in a randomized order. Patterned stimulation trains consisted of 10 sets of 50 Hz triplets, delivered at a frequency between 3 and 7 Hz in 0.2 Hz intervals administered in a randomized order. B Recorded EEG was preprocessed to remove artefacts and identify epochs pre and post each train of stimulation for both rhythmic and patterned paradigms. Percentage change in the power spectrum was calculated and significant changes in response frequency bands were determined from both scalp and source signals using linear mixed models. C The Influence of the DLPFC on all regions was calculated using dynamic network models. Regions that showed a significant change before and after stimulation were determined using a linear mixed model and plotted on source space.