Research Area:  Machine Learning
Any neurological ailment that critically torments the day-to-day survival of patients, reveal its epilogue in any of the alpha, beta, theta or delta spectrum of brain waves. EEG displays the electrical activity of brain and, nowadays it is the most common tool for the diagnosis of neurological maladies. This paper aims at effective de-noising of EEG signals. EEG is obtained by recording the spontaneous electrical activity of the brain over a period of time and it may contain a whole lot of information. This information can be decoded by signal processing methods, but in most cases artifacts interrupt these signals. The recommended approach is based on ICA which is proven better by performance analysis having 96% accuracy and can decompose EEG recording into a number of event related and artifacts related potentials. This study shows that the proposed method significantly enhance the classification accuracy, by effective identification and removal of artifacts.
Keywords:  
Author(s) Name:   K. G. Anjana Lakshmi; S. N. Nissa Surling; O. Sheeba
Journal name:  
Conferrence name:  International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET)
Publisher name:  IEEE
DOI:  10.1109/WiSPNET.2017.8300232
Volume Information:  
Paper Link:   https://ieeexplore.ieee.org/abstract/document/8300232