A Fresh Look at Functional Connectivity Analysis

Abstract

Functional connectivity analysis attempts to detect methods of communication between groups of neurons in the brain that may or may not be directly physically connected. We briefly review current neuroscience methods for functional connectivity analysis, noting the similarities between the current state of the art as described in Miller et. al [4] and the recent work in the field of complex-valued statistics pioneered by Scharf et. al [7] and Picinbono et. al [5]. We apply the techniques of these complex-valued statistics to simulated and natural data, showing that these techniques are able to correctly and accurately detect a common mode of communication known as amplitude/phase coupling in the neuroscience community. This analysis is shown to reinforce the results of the study of in [4], with the added and substantial benefit of being completely automatic and not requiring human intervention to produce results.