AI-assisted real-time supervision for oral myofunctional training for obstructive sleep apnea
Project Description
Obstructive sleep apnea syndrome (OSAS) is a common sleep disorder affecting one billion adults worldwide. The new emerging form of treatment - oral myofunctional therapy is a structured program of neuromuscular re-education targeting the pharyngeal, lingual and perioral musculature. It is designed on the basis that the OSA is caused by recurrent upper airway collapsibility during sleep, by strengthening the upper airway muscles through active exercise and increasing the passive muscle tension can reduce the degree of collapse which relieves the obstruction during sleep. Randomized control trials have demonstrated that a standard 12-week protocol can reduce Apnea-hypopnea Index (AHI) by 50% in adults, alongside improvement in sleep quality and snoring. This form of treatment has been proven to be as effective as surgery without the invasiveness but its effect is dependent on the patient's commitment and compliance. Traditionally the protocol is led by healthcare personnel such as sleep technologists or speech therapists with real-time feedback during clinical sessions followed by a self-practice period after each lesson. Our service is unable to meet the demand due to manpower constraints, with the advancement in technology particularly in facial recognition, we hope to extend the implementation to the clinical field to deliver treatment to patients.
Supervisor
SHEN, Yiwen
Co-Supervisor
WANG, Jiying
Quota
2
Course type
UROP3100
UROP3200
UROP4100
Applicant's Roles
Remote OMT training via smartphone app with real-time feedback: to develop an intelligent - smartphone based AI system capable of real-time OMT supervision that provides automated kinematic analysis, corrective biofeedback and continuous adherence tracking.
Applicant's Learning Objectives
To deliver remote OMT training with real-time feedback, meeting clinical service demands and reducing manpower needs.
Complexity of the project
Challenging