Feedback inhibition control in whole-brain simulation models. Implementation and analysis of a local homeodynamic solution in the Jansen-Rit model

In my thesis, I have introduced an inhibitory plasticity mechanism into the widely-used Jansen Rit (JR) mean field neuronal population model in the context of large-scale simulations of brain network models (BNMs). First, I provided a broad introduction to the field and discussed the components of the whole-brain simulations setup, outlining the motivation and context for the presented work. The proposed dynamic Feedback Inhibition Control (dFIC) addressed the problem of global over-excitation arising from structural connectivities with heavy-tailed weight distributions. I demonstrated the implementation of dFIC in a single node, small network, and whole-brain scales. Single-node analysis resulted in the derivation of the necessary conditions and limitations of dFIC in a single-node uncoupled JR model and their extrapolation to the network case, including the determination of feasible tuning targets. Small network results provided evidence supporting such extrapolation, exploration of dFIC-specific parameters, and demonstrated the effect of noise on dFIC in the JR system. Further, the implementation of dFIC in the whole-brain network highlighted its impact on the characteristics of both the outputs of the JR model. The introduction of advanced fitting methods and the novel application of Poincaré analysis allowed me to establish a connection between the increase in the complexity of the dynamics exhibited by the JR model and the fitting optima.