Neurological abnormality detection from electroencephalography data: a review

摘要

The efficient detection of neurological abnormalities (disorders) is very important in clinical diagnosis for modern medical applications. As stated by the World Health Organization’s (WHO), brain diseases like Alzheimer’s disease, epilepsy, and stroke to headache, infected almost one billion people globally. Electroencephalography (EEG) is the current reference standard for diagnosis of most of the neurological diseases as it is inexpensive, bearable, and non-invasive compared to other tests (e.g. computed tomography, positron emission tomography, mini-mental state examination, and magnetic resonance imaging). Many studies are performed using EEG signals to detect the neurodegenerative abnormalities in the preliminary stage. This paper attempts to provide a comprehensive survey on the recent studies which are made using EEG signals to detect the neurological diseases: Dementia, Mild Cognitive Impairment, Alzheimer, Schizophrenia, and Parkinson. This paper focuses on the following key research questions: (1) what are the key components of EEG signal processing, (2) what algorithms have been used in this processing, (3) which signal processing techniques have received more attention? The paper provides a clear description of the mentioned neuro-diseases along with the relevant studies. Moreover, this study presents all the recent efforts of the methods that are obtained various step of signal data processing including feature extraction and classification phases. Finally, an elaborated comparison of the existing efforts with their drawbacks and performance are reported. This will guide the medical field researchers and technology experts to discover more accurate solutions for neuro-diseases and come up with a neurological abnormality detection framework.