Algorithms and Bounds for Dynamic Causal Modeling of Brain Connectivity

摘要: Recent advances in neurophysiology have led to the development of complex dynamical models that describe connections and causal interactions between different regions brain. These are able accurately mimic event-related potentials observed by EEG/MEG measurement systems, considered be key components for understanding brain functionality. In this paper, we focus on a class nonlinear dynamic (DCM) described set connectivity parameters. practice, DCM parameters inferred using data obtained an EEG or MEG sensor array response certain event stimulus, then used analyze strength direction regions. The usefulness these process will depend how they can estimated, which turn noise, sampling rate, number samples collected, accuracy source localization reconstruction steps, etc. goals paper present several algorithms parameter estimation, derive Cramer-Rao performance bounds estimates, compare against theoretical limits under variety circumstances. influence noise rate explicitly investigated.