Text S1 - Human Motor Cortical Activity Is Selectively Phase-Entrained on Underlying Rhythms

The supplement, Text S1, contains a variety of material that is designed to provide further insight for the more curious reader, further methodology for the more technical reader, and further evidence for the more skeptical reader. It illustrates the interaction between broadband amplitude and rhythm phase using multiple alternate and independent methods of calculation (fures S1, S3, S4). The averaged coupling palette from finger-specific electrodes across many subjects shows that the most robust entrainment in pre-central motor cortex is between 12–20 Hz (Figure S2). The somatotopic representation of individual fingers, and illustrative timecourses (as in Figure 3) are shown in detail for 6 additional subjects (Figures S5, S6, S7, S8, S9, S10). The close correlation between somatotopic representation in broadband and electrocortical stimulation (ECS) is shown for the patients where ECS was performed (Figures S7, S8, S9, S10). We show simple schematics for the relationship between EEG and ECoG for the synchronous (rhythmic and ERP) and asynchronous aspects of the cortical surface potential; a back-of-the-envelope calculation is used to measure the relative contributions of these phenomena (Figure S11). We implement simple heuristics to generate simulated timeseries which have phase entrainment of a 1/f - type signal on an imposed oscillation, illustrating how our analysis techniques extract these phenomena in the expected way (Figure S12). We then apply our analysis techniques to the negative control case of simple 1/f-type colored noise to show that the methods do not self-consistently report phase entrainment (Figure S13). A simple heuristic illustration shows how, and why, the “diagonal bands” in some coupling palettes might exist, in the case where broadband power is elevated at a fixed time lag relative to the peak of a periodic voltage trace (Figure S14). We then illustrate a candidate simple large-scale mechanism for the generation of rhythmic phase-entrainment on local cortical activity (Figure S15). For those readers interested in the preferred phase of entrainment as a function of anatomic region, this is explicitly quantified in Figure S16. The conditional parametric relationship between broadband amplitude, beta (β, 12–20 Hz) amplitude, and magnitude of phase entrainment is further explored and illustrated in Figure S17. Finally, the supplemental material highlights and illustrates phase coherence in further detail, beginning with a simplified illustration of pair-wise phase coherence relative to a reference “seed” site, in Figure S18. Adding to Figures 11 and 12, the pair-wise 12–20 Hz coherence between all sites, and a peri-central index-finger specific site, is shown for the remainder of the subjects, in Figure S19. Figure S20 contrasts the different (yet both gyrally constrained) spatial distributions of coherence for the theta (4–8 Hz) and beta (12–20 Hz) frequency ranges. (PDF)