A3 Refereed book chapter or chapter in a compilation book
Artificial intelligence–based depression detection using EEG signals
Authors: Tokmak, Fadime; Subasi, Abdulhamit
Editors: Subasi, Abdulhamit
Publisher: Academic Press
Publication year: 2024
Book title : Applications of Artificial Intelligence in Healthcare and Biomedicine
Series title: Artificial Intelligence Applications in Healthcare and Medicine
First page : 69
Last page: 93
ISBN: 978-0-443-22308-2
DOI: https://doi.org/10.1016/B978-0-443-22308-2.00007-X
Web address : https://doi.org/10.1016/B978-0-443-22308-2.00007-X
Abstract
Depression is a mental health disorder that significantly impacts individuals' well-being and quality of life. Early and accurate detection of depression is crucial for effective treatment. In this chapter, we applied different AI methods, including convolutional neural networks (CNN) and transfer learning methods for depression detection using EEG signals. This study utilized collected EEG data from 33 patients diagnosed with depression and 28 participants during mental task state. The EEG signals are transformed into continuous wavelet transform images, which capture the temporal and frequency information of brain activity. With 10 different pretrained models, 7 different CNN models were designed and employed to learn discriminative features from the EEG images. The experimental results demonstrated the effectiveness of the AI models for depression detection. The CNN model achieved an accuracy of 95.536% in distinguishing between depressed individuals and the control group. The findings revealed that transfer learning is also useful for depression detection and VGG19 yielding the highest accuracy among the tested architectures. The results highlight the potential of CNN-based approaches and transfer learning approaches for accurate depression detection using EEG signals. The ability of the CNN model to extract meaningful spatial features from EEG signals contributes to its efficacy in discriminating between depressed and nondepressed individuals. These findings have implications for advancing mental health diagnostics and treatment, as the proposed approach offers a promising tool for early detection of depression and intervention planning.
Depression is a mental health disorder that significantly impacts individuals' well-being and quality of life. Early and accurate detection of depression is crucial for effective treatment. In this chapter, we applied different AI methods, including convolutional neural networks (CNN) and transfer learning methods for depression detection using EEG signals. This study utilized collected EEG data from 33 patients diagnosed with depression and 28 participants during mental task state. The EEG signals are transformed into continuous wavelet transform images, which capture the temporal and frequency information of brain activity. With 10 different pretrained models, 7 different CNN models were designed and employed to learn discriminative features from the EEG images. The experimental results demonstrated the effectiveness of the AI models for depression detection. The CNN model achieved an accuracy of 95.536% in distinguishing between depressed individuals and the control group. The findings revealed that transfer learning is also useful for depression detection and VGG19 yielding the highest accuracy among the tested architectures. The results highlight the potential of CNN-based approaches and transfer learning approaches for accurate depression detection using EEG signals. The ability of the CNN model to extract meaningful spatial features from EEG signals contributes to its efficacy in discriminating between depressed and nondepressed individuals. These findings have implications for advancing mental health diagnostics and treatment, as the proposed approach offers a promising tool for early detection of depression and intervention planning.
UTU Research Portal