Pleth-SleepNet: A Unified Network for Simultaneous Sleep Stages and Disorder Classification With PPG Signal

The objective of this project is to accurately detect and classify sleep stages and disorders using plethysmography-based signals, categorized as Healthy, Mild, Moderate, and Severe. By leveraging deep learning models—hybridCNN_GlobalPooling and LightweightAttentionCNN—the project aims to enhance the monitoring and diagnosis of sleep-related health conditions. The primary goal is to develop an automated system that can identify different levels of sleep quality and disorder severity with high precision using image-based signal representations. This system will provide valuable insights for clinical decision-making, enabling early detection and intervention for sleep disorders, thereby improving patient care and health outcomes.