Method for single-trial readiness potential identification, based on singular spectrum analysis

This paper presents a brief description and the advantages of the singular spectrum analysis (SSA). SSA has recently been recommended for analysis of short, noisy time series. The method was applied to single trials of EEG activity at Ct, Fz, C4, Pt and C3, recorded 3.6 s prior to and 1.2 s after the onset of a voluntary motor act. This specific activity is known in its averaged form as readiness potential (RP) and is considered to reflect the preparation of the voluntary movement. However, some physiological investigations require identification of the parameters of single-trial RPs - their onset and successive phases. SSA is based on analysis of the principal components in vector space of delay coordinates of time series. As a result, SSA algorithm decomposes real data records to components reflecting the trend, alpha and beta frequency bands, respectively, whose oscillations appear or disappear in different time instants. The latter were interpreted to distinguish different dynamical stages of the movement preparatory process.