Application for Detection of Sleep Apnea using Machine Learning

Sleep apnea is a disorder characterized by repeated drops in oxygen saturation and fluctuations in heart rate during sleep. Clinical diagnosis through polysomnography is accurate but expensive, time-consuming, and not easily accessible for routine screening. This project presents a portable and affordable sleep apnea monitoring system that uses a pulse oximeter sensor interfaced with a microcontroller to continuously measure SpO₂ and heart rate. The microcontroller collects real-time physiological data, processes the signals, and identifies potential apnea events based on oxygen desaturation patterns and abnormal pulse-rate variations. The system stores the readings for later evaluation and provides a simple method for preliminary detection without the need for complex clinical equipment. Experimental results indicate that monitoring these two key parameters using a pulse oximeter is sufficient to capture physiological changes commonly associated with sleep apnea. The proposed design offers a practical and low-cost alternative for home-based screening and early identification of sleep-related breathing disorders.