Sleep Stage Recognition from EEG Signals via 1D Convolutional Networks

Sleep is essential for both physical and mental health and greatly improves the accurate diagnosis of sleep disorders. A common disorder is sleep apnea (SA), which occurs while breathing in while sleeping. This can lead to brain and heart problems and, in severe cases, lead to stroke or cardiovascular disease. Additionally, SAs increase the risk of accidents, especially on day to day. Given its complexity, early detection and treatment of SA is critical. To address this, interest in automating SA awareness has increased. With a large amount of medical data, extracting meaningful information requires rapid and effective analysis. Classifying sleep levels traditionally requires specialized knowledge and is complex and time-consuming. This means you need an automated solution that uses EEG data. CNN is effective in learning patterns in EEG data and shows strong accuracy in identifying sleep levels. Future research aims to improve the scalability of these methods, integrate data from several sources, and examine the real applications of portable devices for more advanced sleep monitoring.