A Review on Exploration of EEG-Based Mental Illness Detection Tools and Techniques

The difficulties with road accident rates today rank among the top concerns for health and social policy in nations across the continents. Now-a-days, electroencephalogram (EEG) is considered to be one of the most investigating tools for diagnosis of diseases. Some of the common mental illnesses are Alzheimer’s disease schizophrenia, obsessive compulsive disorder (OSD), anxiety, depression, etc. So, proper diagnosis is needed to be done as mental illness is most dangerous illness that deteriorates not only individual’s life but also affects entire family and society. In this paper, a review is presented, and framework is designed to merge the efficacy of machine learning and for mental illness. Paper presents a review on methods to identify the healthy control (HC), high mental impairment (HMI), and mild mental impairment (MMI) using EEG signal. Paper has presented a brief analytical review on EEG signals and their processing steps. The paper investigated the EEG signal processing techniques and application of machine learning (ML) for mental health diagnosis. For preprocessing, it was observed that Independent Component Analysis (ICA) or Transient Artifact Reduction Algorithm (TARA) was most efficient methods, whereas hybrid feature extraction techniques result in better outcomes. And for classification, ensemble learning or deep learning showed better performance for classifying mental state. Therefore, this paper motivates the researchers for development of future EEG diagnostic tools.