A Deep Learning Framework for Comprehensive Sleep Analysis and Quality Tracking

Sleep influences people's lives but remains mysterious in many ways.As we sleep, we go through cycles of ‘deep’, ‘light’, and ‘rapid eye movement’ (REM) sleep. Most sleep trackers calculate how much time users likely spend awake versus asleep using a wearable gadget, such as a smartwatch. Some tools use changes in heart rate while you sleep to determine how long you spend in each sleep cycle. Despite their popularity, only a few studies have investigated how accurate sleep devices are. For the same, A non-obstructive and credible Sleep analysis is essential to track sleep consistently over time to get a clear picture of sleep patterns and identify any potential issues. The paper proposes a novel approach to identify sleep-associated disorders and analyze an individual's sleep cycles through different parameters such as ECG, blood oxygen levels, body movements, and other physical parameters. The proposed system will be capable of addressing the limitations of the polysomnography (PSG)test, including its complex methodology, discomfort for the user, and restriction of movement.