Recognition of Sleep Apnea and Hypopnea Events from Single Channel Respiration Signal for Healthcare Application

In recent years, sleep apnea has emerged as a major contributor to the epidemic of sleep disturbances that has emerged in modern society. Due to the high price, unpleasant experience, and lack of available sleep clinics associated with sleep apnea diagnostics, there is a need for home-based health monitoring systems that are both affordable and convenient. Recent years have seen a lot of research into methods for detecting apnea and hypopnea, but less attention has been paid to classifying the many types of events. Since obstructive and central apnea require different treatment strategies, identifying the specific type of apnea is just as crucial as spotting the condition itself. The work makes use of both an oronasal airflow signal and an abdominal effort signal. The intricate link between these two signals throughout various stages of apnea has been studied using a simulated lung airflow model. This idea has been refined in clinical trials to include the ability to identify the nature and timing of events in a near-real-time setting. The average accuracy of detecting hypopnea and apnea during online tests was 92.4% and 95.3%, respectively, which is pretty satisfactory when equated with Polysomnography (PSG) based related directories on undetected participants.