Downloads: 2 | Views: 46 | Weekly Hits: ⮙1 | Monthly Hits: ⮙2
Research Paper | Computer Science and Information Technology | United States of America | Volume 13 Issue 1, January 2024 | Popularity: 5.1 / 10
AI-Driven Sleep Disorder Classification Using EEG Data
Omkar Reddy Polu
Abstract: Sleep disorder has a huge impact on the human health, includes cognitive decline, metabolic dysfunction, and cardiovascular diseases. Manual analysis of EEG signals in the case of sleep disorders is traditionally the basis of diagnosis, which is a slow and error prone process. On this regard, we propose a sleep disorder classification framework based on AI driven deep learning models to automate and improve the sleep stage analysis performance. We develop a hybrid deep learning model that combines Convolutional Neural Networks (CNNs) to extract spatial features and transformers to learn long range temporal dependencies in the EEG data. Furthermore, we apply transfer learning on pre-trained EEG models to improve the classification performance with lower computational overhead. We further embed Explainable AI (XAI) techniques like SHAP and Grad-CAM in order to make the model more interpretable and enable clinicians to understand how the model makes its decisions. To further improve classification accuracy, we fuse multi-modal physiological signals, i.e. HRV and SpO?. Additionally, federated learning keeps the model training across multiple hospitals private by not sharing patient data that is sensitive. We apply extensive experiments on benchmark EEG datasets, which show that our model significantly outperforms state-of-the-art methods in terms of both accuracy and generalization across various populations. The solution proposed by this research is robust, interpreted, and privacy aware for real time automated sleep disorder diagnosis to support better clinical decision making.
Keywords: Sleep Disorder Classification, EEG-Based Deep Learning, Transformer Neural Networks, Multi-Modal Data Fusion, Federated Learning in Healthcare
Edition: Volume 13 Issue 1, January 2024
Pages: 1844 - 1849
DOI: https://www.doi.org/10.21275/SR24011114422
Please Disable the Pop-Up Blocker of Web Browser
Verification Code will appear in 2 Seconds ... Wait