Artifact Processing of Epileptic EEG Signals: An Overview of Different Types of Artifacts

The electroencephalogram (EEG) signal contains a plethora of information regarding a human brain and thus plays a very important role when it comes to the diagnosis of various neurological disorders. The quantitative analysis of epileptic EEG can provide a neurologist additional and very important information related to epileptic seizures both in terms of identification and prediction. However, EEG data is always contaminated with different types of artifacts and as such any kind of analysis carried out on such raw data would lead to spurious results. Besides artifact identification and removal techniques, this system should include high quality recording of EEG and minimization of artifacts coming from different sources. Thus any effort towards actualization of an automatic analysis system for EEG in general and epileptic EEG in particular will require an efficient, reliable and robust artifact processing mechanism. The distinction between cerebral electrical activity and artifacts is crucial for actual interpretation and diagnosis. This paper presents various types of artifacts in the epileptic EEG and review of different methods used for identification and removal of these artifacts.