Why the study?
Depression is a prevalent, debilitating illness requiring patient-friendly, cost-effective early diagnosis based on objective indicators like EEG signals.
Do machine learning and deep learning frameworks utilizing EEG signals improve the diagnostic accuracy of depression in patients with Major Depressive Disorder?
Population
34 patients with Major Depressive Disorder and 30 healthy subjects
Comparison
1DCNN vs SVM vs LR across TASK, Eye Close, and Eye Open EEG signals
Design
Diagnostic classification and comparative machine learning study
Key result
EEG-derived TASK signals analyzed with a 1D Convolutional Neural Network achieved the highest accuracy of 90.21% for diagnosing depression (p < 0.05).
Authors
Loading...
May support EEG-ML for MDD diagnosis; leaves open validation before clinical adoption.
Observational (n=64)
Do machine learning and deep learning frameworks utilizing EEG signals improve the diagnostic accuracy of depression in patients with Major Depressive Disorder?
p-value: p=< 0.05
A 1D Convolutional Neural Network utilizing task-based EEG signals achieved 90.21% accuracy in diagnosing major depressive disorder.
Nitin Ahire (2025) conducted an observational in Major Depressive Disorder (n=64). EEG-based machine learning and deep learning classifiers (1DCNN, SVM, LR) vs. Healthy controls was evaluated on Classification accuracy for diagnosing depression using TASK signals (p=< 0.05). EEG-derived TASK signals analyzed with a 1D Convolutional Neural Network achieved the highest accuracy of 90.21% for diagnosing depression (p < 0.05).