The Study Of Motor Imagery EEG Based On EEG Microstate Analysis

Electroencephalogram(EEG)is the overall electrical physiological activity of nerve cells reflected on the cerebral cortex and scalp,it contains a lot of information related with diseases and physiology.Because the correlation between person's state of consciousness activity and brain electrical signal,via the classification of the eeg signals,we can identify different brain state of consciousness activity.Thus forming a new type of human-comp uter interaction(that is Brain-Computer Interfacetechnique)making person communicate with the external environment through computers and other electronic equipment output control signals without participation of peripheral nervous system and muscle tissue.In recent years,brain-computer interface based on EEG becomes the focus in the many fields such as brain science,neuroscience,artificial intelligence,etc.The classification on motor imagery EEG is an important branch in the field of Brain-machine interface.This paper studied the differences in parameters of EEG Microstates between left and right hand motor imagery EEG via EEG Microstate analysis.Then Modeled and classified these two types of EEG by using hidden markov model(HMM).The main contents are as follows:1.Motor imager EEG signal preprocessingEEG signals,as a kind of bioelectricity physiological signals,inevitably interferenced by various noise such as electrooculogram,myoelectricity and electrocardio,etc.In this paper,independent component analysis was used to remove these artifacts to facilitate subsequent processing and analysis.2.EEG Microstate analysis of Motor Imagery EEGEEG Microstate analysis was used to study left and right hand motor imagery EEG of nine subjects.Four kinds of typical EEG Microstates was obtained from Multi-channel eeg signals.And compared the differences in Microstate time series parameters between the two kinds of signals.Parameter characteristics of the Microstate 3 and 4 have signi ficant differences between the two kinds of imagine tasks: During the left hand imagination movement,parameters of The Microstate 4 are increased while parameters of state 3 are reduced.During the right hand imagination movement the situation is reversed.3.Classification of the tyo types of EEG based on EEG Microstate parametersHidden Markov Models was used for time series efficient processing capability to model the two types of EEG Microstate parameters and classify them.The method achieved a highe r classification accuracy(the highest result up to 98% while the average was 89.44%).Higher than that of other methods on the average level.