Change-point Analysis of Single-trial EEG of Auditory and Visual Oddball Tasks

the brain rhythms defined by specific frequency bands of electroencephalography (EEG) signal are generally thought to represent diverse cognitive processes which wax and wane with sub-second resolutions at different parts of the cortex. Furthermore, single-trial analysis of EEG is believed to show more realistic pictures of ongoing and event-related activities of the brain with millisecond resolution. Here, we present a nonparametric multiple change-point detection and estimation method for analysis of single-trial EEG recorded during auditory and visual oddball tasks. In a simple attention task like oddball, the frontal cortex of the brain is responsible for distinguishing and responding appropriately to target and standard events. With sub-second resolution at the frontal cortex, we show that the alpha-band activity changes according to "inhibition timing" hypothesis and the beta-band activity is in line with "maintaining the status quo" hypothesis.