Sleep is a physical and mental state that plays an important role in the maintenance of health and well-being. It is responsible, for example, for the functions of memory consolidation and metabolism regulation, which, when absent, can have mental consequences and severe physiological disorders, called sleep disorders. These disorders are syndromes that affect normal sleep, negatively affecting health, as in the case of Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS), which is related to the momentary stopping of breathing at night.
Snoring is one of the main indicators of the presence of sleep disorders related to respiratory causes. By means of the detection and analysis of these indicators we can extract characteristics that allow the identification of respiratory disorders, which constitutes a possibility of previous detection of these, accelerating the diagnosis and access to treatment.
The goal of this research project is to develop a mobile application for monitoring the user's night of sleep in order to aid in the diagnosis and treatment of sleep disorders. The application called MySleep performs the recording and transmission of audio captured at night to a server responsible for analyzing this signal and extracting respiratory events from snoring and OSAHS in order to estimate the rate of apnea-hypopnea.
The captured audio signal is being analyzed through digital signal processing, artifical intelligence and machine learning techniques. A first version of the Android application is now available for testing, with audio recording and streaming features, and has guidance screens for signal capture and user instruction on care to be taken before recording starts. The next steps will be to extract and display quality metrics and sleep disturbances visually within the application and make sleep metrics available to hospitals and other interested parties through a web portal for follow-up and clinical decision support.