Analysis of multichannel biomedical data

Abstract

Nowadays, there is a common practice in biomedical research to perform multiple time series recordings. In the first part of this paper, basic information about analysis of such multichannel biomedical data is given. A short overview of important differences between single-channel, two-channel and multichannel data sets is presented and various coherence functions are reported. Causal relations between channels are investigated by means of the Directed Transfer Function (DTF) and its dynamic version, the Short-Time Directed Transfer Function (SDTF). The introduced formalism was used to analyze 12-channel human electrocorticogram (ECoG) records. Preliminary results of a study of causal dependence in beta and gamma frequency bands in two patients performing a motor task are reported. Specific characteristics in activity propagation consistent for both subjects for different rhythms were found.