[Paper Review] EEG based Major Depressive disorder and Bipolar disorder detection using Neural Networks: A review

Sana Yasin, Syed Asad Hussain+4 more|arXiv (Cornell University)|Sep 28, 2020

Functional Brain Connectivity StudiesNeuroscience192 references203 citations

TL;DR

A comprehensive survey of neural network approaches using EEG for detecting Major Depressive Disorder and Bipolar Disorder, summarizing architectures, biomarkers, datasets, and recommendations.

ABSTRACT

Mental disorders represent critical public health challenges as they are leading contributors to the global burden of disease and intensely influence social and financial welfare of individuals. The present comprehensive review concentrate on the two mental disorders: Major depressive Disorder (MDD) and Bipolar Disorder (BD) with noteworthy publications during the last ten years. There is a big need nowadays for phenotypic characterization of psychiatric disorders with biomarkers. Electroencephalography (EEG) signals could offer a rich signature for MDD and BD and then they could improve understanding of pathophysiological mechanisms underling these mental disorders. In this review, we focus on the literature works adopting neural networks fed by EEG signals. Among those studies using EEG and neural networks, we have discussed a variety of EEG based protocols, biomarkers and public datasets for depression and bipolar disorder detection. We conclude with a discussion and valuable recommendations that will help to improve the reliability of developed models and for more accurate and more deterministic computational intelligence based systems in psychiatry. This review will prove to be a structured and valuable initial point for the researchers working on depression and bipolar disorders recognition by using EEG signals.

Motivation & Objective

• Survey and synthesize literature on EEG-based MDD and BD detection using neural networks from 2010 to 2020.
• Summarize EEG protocols, features, biomarkers, and public datasets used in NN-based depression and bipolar disorder recognition.
• Compare shallow and deep neural network approaches in this domain.
• Provide guidelines and recommendations to improve reliability and determinism of models in psychiatry.

Proposed method

• Systematic literature search across multiple databases using PRISMA principles.
• Inclusion criteria focusing on EEG-based depression assessment with artificial neural networks.
• Extraction and synthesis of study designs, NN architectures, EEG features, datasets, and reported outcomes.
• Discussion of limitations, challenges, and future research directions in EEG/NN-based psychiatry.
• Consideration of clinical relevance and potential biomarkers linked to EEG signals.

Experimental results

Research questions

• RQ1 What neural network architectures (shallow vs. deep) have been applied to EEG-based MDD and BD detection?
• RQ2 Which EEG protocols, features, and biomarkers are most commonly used in NN-based depression and bipolar disorder recognition?
• RQ3 What public datasets are available for EEG-based MDD and BD research and how are they utilized?
• RQ4 What are the main limitations of current EEG/NN approaches and what recommendations improve reliability and determinism?

Key findings

• Provides a comprehensive survey of both shallow and deep neural networks used with EEG for MDD and BD detection.
• Discusses EEG biomarkers, feature extraction approaches, and experimental protocols used in the literature.
• Summarizes available EEG-based datasets and their role in model development and evaluation.
• Highlights methodological challenges, limitations, and guidance for improving model reliability in psychiatric applications.
• Positions the review as a structured starting point for researchers investigating depression and bipolar disorder recognition with EEG signals.