Estimation of resting state effective connectivity in epilepsy using direct-directed transfer function

There has been an increasing demand among neuroscientists to understand the complex network of functionally connected neural assemblies in the human brain. For this purpose, computational EEG research is widely used by researchers due to its remarkable advantage in providing high temporal resolution, and ease of analysis across different frequency bands. Here we analyzed Electrocorticographic (ECoG) signals of electrodes placed on frontal-parietal neocortex brain region of 8 pediatric epileptic patients. In order to evaluate the directed causal relationship among different brain regions, we employed a Granger causality based multivariate connectivity estimator named direct Directed Transfer Function (dDTF) to identify signal propagations among the selected set of electrode in the frequency range 1-50Hz. A consistent network pattern emerged that was unique to each patient. The fidelity of such dDTF-derived connectivity patterns can support a clearer understanding of effective connectivity in epileptic networks.