Dynamic Causal Modelling (DCM) for M/EEG

In this introductory talk, I explain generative modelling of M/EEG using Dynamic Causal Modelling (DCM). I present the main principles of DCM and describe the extended 3-population Jansen-Rit neural mass model which constitutes the DCM-ERP model for event-related potentials and cross-spectral densities. I finish with a summary of the key steps involved in a DCM analysis and possible investigations on inverted DCM models. It would be advisable to watch at 1.5x. Sections: 00:00 Introduction 00:26 Dynamic Causal Modeling 03:12 Generative Modeling with DCM 08:56 Types of DCM 12:50 Neural Mass Models for M/EEG 17:00 Modeling the Neocortex 19:43 The Jansen-Rit Model 33:25 Dynamics due to E-I balance 35:18 Connecting cortical columns: Extrinsic connections 37:26 Extrinsic connections: An example architecture 38:52 The extended Jansen-Rit Model or the 'ERP' model 41:55 Neuronal models in DCM 43:16 3-population models: Variations of Jansen-Rit/ERP 46:41 4-population models: The Canonical Micro-Circuit or CMC 47-51 Summary of neuronal models 49:25 Observation models 51:36 DCM analysis pipeline 53:24 Model inversion 55:18 Investigating with DCM: Hypothesis testing 56:26 Summary