An Update to the Boston EEG Automated Processing Pipeline Kalman Detrend Step and an Examination of Mu Suppression

Abstract

When analyzing electroencephalography (EEG) data, noise must be removed from the observed signal in order to parse out the "true" EEG signal. The Kalman detrend step within the Boston EEG Automated Processing Pipeline (BEAPP) removes low-frequency noise from the observed EEG signal. The first part of this thesis develops an update to the Kalman detrend step of BEAPP. This update uses expectation maximization to allow researchers to more accurately and uniquely remove this noise from an observed EEG signal. More rigorous implementation must be done before this new detrend step can be added to BEAPP and used to clean EEG data. The second part of this thesis uses the currently implemented Kalman detrend step within BEAPP to examine mu suppression in individuals with autism spectrum disorder (ASD). In typically developing individuals, females have been shown to display more mu suppression than males when observing human actions. The gender differences in ASD are not well understood and this gender pattern in mu suppression has not been explored in individuals with ASD. Mu suppresion was analyzed by comparing the average mu power over the sensorimotor cortex in an eyes closed resting state and when observing the Biological Motion paradigm. The results found no differences in mu suppression between males with ASD and females with ASD.