Developing a Technique for Characterization of Upper Airway and Screening of Obstructive Sleep Apnea Using Tracheal Breathing sounds
Obstructive sleep apnea (OSA) is a prevalent yet underdiagnosed health problem. Assessment of OSA is currently based on sleep studies that are time-consuming and expensive. This proposal presents three research projects/points to apply machine learning techniques and statistical tests on tracheal breathing sounds (TBS) signals for OSA screening. We will investigate the pathology of the OSA using TBS analysis during wakefulness, sleep, and in the transition from wakefulness to sleep, compare various techniques for feature selection and classification, and finally enhance the current OSA screening algorithms. The main expected outcomes of this work will be finding the TBS characteristics that reveal the structural and physiological changes of UA in relation to OSA in a detailed but straightforward manner. Also, providing a non-time-consuming and less expensive method to stratify the severity of OSA patients in a fast but more precise way.