Induced synchronisation by endogenous noise modulation in finite-size random neural networks: a stochastic mean-field study
Résumé
Event-related synchronisation and desynchronisation (ERS/ERD) are well-known features found experimentally in brain signals during cognitive tasks. Their understanding promises to have much better insights into neural information processes in cognition. Under the hypothesis that neural information affects the endogenous neural noise level in populations, we propose to employ a stochastic mean-field model to explain ERS/ERD in the γ−frequency range. The work extends previous mean-field studies by deriving novel effects from finite network size. Moreover, numerical simulations of ERS/ERD and their analytical explanation by the mean-field model suggests several endogenous noise modulation schemes which may modulate the system's synchronisation. Endogenous neuroelectric fluctuations are assumed to contribute or reflect information processing in the brain. We hypothesize that these fluctuations may induce event-related synchronisation and desynchronisation (ERS/ERD), which are spectral power enhancements and reductions, respectively, in certain frequency bands. A stochastic network model considers such endogenous noise and exhibits ERS/ERD in the γ−frequency range dependent on the endogenous noise level. Mathematical analysis and numerical simulations reveal the underlying mechanism called Additive-Noise Induced System Evolution (ANISE) in line with experimental results. The work proposes an auspicious perspective to explain ERS/ERD in various frequency bands to reveal underlying mechanisms of brain information processing.
Domaines
Sciences cognitivesOrigine | Fichiers produits par l'(les) auteur(s) |
---|