Quantitative and qualitative discrimination of task periods from non-task periods by recurrence plots

Abstract

Recurrence plots are useful tools for visualizing the inner structure of a time series, and they have been applied to physiological data to examine brain activity. However, these studies have focused on qualitative analysis of the data. In this study, we used unthresholded recurrence plots of near-infrared spectroscopy (NIRS) data to detect the difference in brain activity between task periods and non-task periods. Histograms derived from the recurrence plots showed a statistical difference between the periods. Throughout the pre-task period, task period, and post-task period, the histogram kept its shape statistically and shifted horizontally according to the period. Therefore, the changes in dynamical systems describing brain activity were observed as changes in histograms derived from NIRS data.