Suppression of Artifacts from EEG Recordings Using Computational Intelligence

Brain–computer interface system will be useful for physically disabled people to analyze and diagnose different health problems. The signal processing module is a major part of the brain–computer interface system. It is divided into four sub-modules notably, pre-processing, feature extraction, feature selection, and classification. EEG captures small amounts of brain activity and is a well-known signal acquisition method, due to its good temporal resolution, cheap cost, and no significant safety concerns. The objective of the EEG-based brain–computer interface system is to extract and translate brain activity into command signals which helps physically disabled people. The EEG recordings are contaminated by a variety of noises generated from different sources. Among these, the eye blinks have the greatest influence on EEG signals because of their high amplitude. This chapter provides a detailed review of the basic principles of various denoising methods, which also succinctly presents a few of the pioneer’s efforts. Further, the comparative analysis is carried out using EMD, AVMD, SWT, and VME-DWT methods for filtering eye blink artifacts. The VME-DWT method is found to perform better than the SWT, AVMD, and EMD methods in terms of signal information retention, which perfectly encapsulates the relevance of our quantitative study. Computational intelligence develops a new approach for identifying and analyzing discriminating characteristics in signals. An EEG-based brain–computer interface system should use computational intelligence to reduce the noises from EEG data proficiently.