Biophysical network modeling of brain dynamics

2017-2022 .
GitHub

Summary

The human brain is highly complex, and even a simple task involves several factors, e.g., attention, emotion, memory, learning, decision making, physiological complexity of the brain itself, and the environment. Hence, it is currently impossible to create a single computational model to account for all aspects of the brain. In fact, it is impractical to model even a single factor completely. However, the goal of computational modeling is not necessarily to build absolutely correct theories. Some of the most successful models are based on approximate and relatively crude underlying assumptions. Rather, the goal is to enable progress in understanding a particular complex phenomenon. The complex phenomena is simulated based on an initial formulation, of a theory, generating novel hypothesis from this simulation, collecting new data based on these predictions, and then refining the model (and the underlying theory itself). The process thereby implements the classic cycle of theory development, testing, and revision to advance the field. We actively use computational modeling of neuroimaging data (EEG/fMRI/fNIRS) to not just tie together neural activity and behavioral findings, but also advance our understanding of brain processes by providing formulations and testable predictions about the underlying mechanisms.

Presentations/Papers

  1. Saggar, M., Zanesco A.P., King, B., Bridwell, D.A., MacLean, K.A., Aichele, S.R., Jacobs, T.L., Wallace, B.A., Saron C.D. & Miikkulainen R (2015) Mean-field thalamocortical modeling of longitudinal EEG acquired during intensive meditation training. NeuroImage

  2. Saggar, M., Miikkulainen R. & Schnyer D. M. (2010) Behavioral, neuroimaging, and computational evidence for perceptual caching in repetition priming. Brain Research

  3. Saggar, M., Markman A.B., Maddox W.T., & Miikkulainen R. (2007) A Computational Model of the Motivation-learning Interface. Proceedings of the Cognitive Science Society

  4. Saggar, M., Mericli T., Andoni S. & Miikkulainen R. (2007) System Identification for the Hodgkin-Huxley Model using Artificial Neural Networks. IEEE International Joint Conference on Neural Networks (IJCNN)

Collaborators

Funding