Figure 1

Methodological workflow and features extraction. TMS-induced electric field was modelled with SimNIBS ¹⁶. Fiber tractography and resting state functional connectivity (rs-FC) were computed from the point with the maximal E-field. As for rs-FC analysis, the final maps were obtained z-transforming, thresholding and FDR-correcting the raw connectivity profile. As for tractography, the reconstructed tract was used to threshold the grey matter map where only the voxels reached by a streamline were retailed. The final rs-FC map and the tract-thresholded gray matter were used to extract the source reconstructed EEG timeseries in order to establish which between functional and structural connectivity mostly predict the TMS-induced signal propagation. The EEG signal was projected at source level using the minimum norm estimation (MNE) method with current density and constraining source dipoles to the cortical surface. The raw time series were first rectified¹⁷ and then a baseline bootstrapping procedure¹⁸ was applied. Then, 1000 permutation t-test were performed in which the surrogated DWI vs rs-FC difference was computed after each iteration and statistically compared with the real difference¹⁹. Finally, the cluster threshold was determined as the 95th percentile of the cluster’s surrogate distribution and the area under the curve (AUC) of the significant clusters was extracted within the 350 ms post-TMS period, with positive and negative values indicating that the TMS-EEG signal propagated more on DWI or fMRI projection map, respectively. For further details on TMS-induced electric field modelling please refer to Supplementary Materials and Methods. For a visualization of the individual white matter bundle of both DAN and DMN please see Supplementary Figs. S2 and S3. No software was used to generate this figure.