EEG FREQUENCY BANDS AND FUNCTIONAL CONNECTIVITY DYNAMICS EMERGE FROM A 40HZ BRAIN MODEL OF PHASE OSCILLATORS

Biophysical, connectome-based models of brain networks demonstrate meta-stability dynamics, this is the arise of multiple states from the same set of parameters. In particular, networks of delayed coupled oscillators at 40Hz, show an alpha band spectrum similar to that observed in the EEG. Further, the dynamics of the network show other oscillatory modes tuning the coupling parameters and the delays. The spectrum peak reduces its frequency in proportion to the increase of the global coupling strength, and the increase of the mean delay between connections. We used a network of delayed Kuramoto phase oscillators at 40Hz to show that multiple oscillatory modes can emerge from the same coupling and the delay parameters. In order to select the global coupling and mean delay parameter, we used spectral entropy to identify the maximum meta-stability. Then, we analyzed the time series of the simulation to identify the dominant frequency peaks and their corresponding network as a function of time. We found that using a global coupling value of 4 and a mean delay of 21 ms, using the AAL90 connectome, we get the maximum meta-stability with four dominant peaks at 13, 15, 29.5, and 43 Hz. The found peaks correspond with alpha, low beta, high beta, and gamma EEG bands. We also found nine nodes, related to vision brain regions, with significant fractional occupancy in the alpha band. But they alternated the occupancy in sub-networks formed by two to five nodes. In addition, we found ten nodes of olfactory brain regions with occupancy in the low beta band; ten nodes in decision/motivation-related brain areas with occupancy in the high beta band; and eleven nodes, related to the ventral default mode network, with fractional occupancy in the gamma band. Interestingly, fifty-one nodes do not have significant occupancy in any of the four frequency peaks.