To facilitate the study of causal dynamics of large-scale brain networks, we propose to develop, validate, and commercialize innovative software tools for measuring neurophysiological effective connectivity in humans. These tools potentially may be used to provide neurologists with a reliable diagnostic for locating epilepsy seizure foci, and to provide systems and cognitive neuroscientists with new techniques for assessing neural information transfer in the human brain. During Phase I, analysis methods were developed to estimate both stationary and time-varying effective connectivity among selected brain regions. These methods compute time-lagged causal information between time series which represent states of brain activity in selected regions of interest, estimated from scalp EEG. Causal information, as distinct from predictive information, is approximated by discounting identified non-causal confounds. We developed both linear and nonlinear measures, together with associated tests of statistical significance. The approach was applied successfully to (a) simulated data, (b) resting EEG, (c) cognitive event-related EEG data, and (d) ictal onset scalp EEG. In Phase II, we will design and develop an effective connectivity software toolset for research use by cognitive and clinical neurophysiologists. We will validate causal information analysis in two stages: first, by comparing the results of intracranial EEG analysis against known effective connectivities obtained via cortical electrostimulation recordings; and second, by comparing connectivity analyses of scalp versus intracranial EEG. Causal information analysis tools will be used to test connectivity hypotheses in a cognitive neuroscience application, and to evaluate potential clinical utility in epilepsy. Throughout, the new measures will be compared with traditional measures for assessing neurophysiological functional connectivity, ? ? ?
