Artifact Removal from Sleep Hipnograms using Complete Ensemble Empirical Decomposition with Adaptive Noise

Abstract

Objectives: Many sleep related disorders can hamper the performance of an individual and there is an absolute need to research upon such disorders. Methods /Statistical Analysis: The neuro physiologists record an electrical activity of the brain through Electroencephalogram (EEG), since the emphasis of study is sleep. Artifacts are the recorded activities which are not the records of cerebral origin. As the EEG record itself runs for longer periods, it suffers from some artifacts. In this article Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMD-AN) method is proposed for removing the artifacts from the contaminated sleep EEG. Findings: In this article we consider Index of Orthogonality (IORT), Percentage Error in Energy (PEE) and Number of sifting iterations to assess the performance of CEEMD-AN method over the earlier methods are explored. Application/Improvement: The artifact free data gives best results for neuro physiologists to identify the disorder. Keywords: Artifacts, CEEMD-AN, EEG Waves, Noise Adaptive Data Analysis, Sleep EEG