A Review on EEG Based Stress Monitoring System Using Deep Learning Approach

Abstract

Stress is one of prime factor that results into various diseases such as cardiovascular diseases, neurodegenerative diseases and many more. It adversely affects the health of the human resulting into several diseases such as Heart attacks, brain stroke, cerebral palsy etc. To overcome this mental disorder, monitoring stress is strictly essential for prevention and its clinical intervention. In present paper, a system for monitoring mental stress is proposed using electro-physiological signal, Electroencephalogram. It measures as well as records the electrical activities in brain and can be analyzed further to investigate different mental disorders by processing it. The present paper presents a novel system for monitoring stress using deep learning approach. EEG signals are analyzed in time – frequency domain and further a method for classifying it into different classes is proposed using Convolutional Neural Networks. EEG signals are analyzed in time frequency domain by applying Discrete Wavelet transform. Therefore, the application of deep learning algorithms in clinical assessment provides a benchmark for examining various neurological disorders and a highly reliable system can be further employed to make significant contributions in this field.