Phase variants of the common spatial patterns method

Abstract

Numerous studies on EEG data indicate that various brain processes are characterized by phase relationships between different regions of the brain. The development of analytic techniques that can provide a better detection of these phase relationships can lead to a better understanding of such brain processes. In this chapter two variants of the common spatial patterns (CSP) method which are designed to capture such phase relationships, namely, the “phase synchronization”-based CSP (P-CSP) algorithm and the analytic CSP (ACSP) algorithm, are presented. The P-CSP and ACSP methods are analyzed and tested on real EEG data. The nature of the results of the two methods is then discussed, highlighting the differences between them.