Short-time Fourier transform and embedding method for recurrence quantification analysis of EEG time series

Abstract Electroencephalography (EEG) allows recording of cortical activity at high temporal resolution. Creating features useful for the analysis EEG can be challenging. Here we introduce a new method pre-processing time-series resting state and binary task classification using recurrence quantification (RQA) compare it with existing state-of-the-art approach based on signal embedding. To reveal patterns that unfold brain dynamics, present pipeline does not rely selection embedding parameters RQA. Instead signals directly, Short-term Fourier transform (STFT) is used to generate time-series, power spectra from sliding, overlapping windows. Recurrence plots are created in standard way embedded signals, STFT vectors. The efficiency RQA extracted such compared segments correspond open closed eye conditions. In contrast common approaches analysis, no filtering into separate frequency bands was needed. Differences between two representations illustrated histograms UMAP plots. Classification results 95.9% level were obtained selected less than 10 electrodes.