Fig. 1

MEG data and pipeline for obtaining individual brain maps of natural frequencies. (a) Within-session data. A single session from each participant was divided into two parts and pre-processed separately. (b) Between-session data. The two MEG sessions recorded at different times for each participant were pre-processed separately. The following panels (c–e) illustrate the pipeline for mapping the brain’s natural frequencies. (c) Source reconstruction and frequency analysis of each part/session. (d) Whole-group k-means clustering was conducted on 150 spectra per voxel and participant from Part/Session 1. (e) Brain mapping of single-subject natural frequencies was derived separately for Part/Session 1 and Part/Session 2. First, single-subject power spectra obtained from one instance of the data were classified into the whole-group clusters. Next, the peak of the most typical power spectrum defined the natural frequency of each voxel, which were smoothed across neighboring voxels. Final brain maps for Part/Session 1 and Part/Session 2 were composed of single vectors of 1925 natural frequencies, i.e., one value per voxel. Note: all illustrations are based on real data.