Sleep Stages Classification Based on Diverse Resting-State Multi-Channel EEG Using Deep Learning

Distinguishing sleep stages is crucial for diagnosing sleep disorders and enhancing overall health. EEG is frequently utilized for studying sleep phases because of its non-intrusive characteristics and precise tracking of brain function through recording brain activity. This research offers a thorough assessment of using EEG to detect sleep stages, using a range of machine learning methods like neural networks (NN), and deep neural networks (DNN). By analyzing EEG signals and applying techniques for selecting important characteristics and refining the process, the different stages of sleep can be classified as wakefulness, non-REM stages, and REM sleep. The models are developed using EEG data from diverse resting state datasets to accurately classify sleep stages based on dominant brain waves including delta, theta, alpha, and beta. Moreover, this study investigates the use of non-linear methods for feature extraction to enhance precision in detection, while also emphasizing the effectiveness of deep learning models for improving classification accuracy. The suggested techniques show encouraging outcomes in automatically categorizing sleep stages, which could have a major impact on improving sleep quality and identifying sleep disorders.