Vision Analysis in Detecting Abnormal Breathing Activity in application to Diagnosis of Obs.pdf

Vision Analysis in Detecting Abnormal Breathing Activity in application to Diagnosis of Obstructive Sleep Apnoea Ching Wei Wang, Amr Ahmed and Andrew Hunter Vision and Artificial Intelligence Group, Department of Computing Informatics, University of Lincoln Brayford Pool, Lincoln LN6 7TS, United Kingdom {cweiwang | aahmed | ahunter}@lincoln.ac.uk Abstract –Recognizing abnormal breathing activity from body movement is a challenging task in machine vision. In this paper, we present a non-intrusive automatic video monitoring technique for detecting abnormal breathing activities and assisting in diagnosis of obstructive sleep apnoea. The proposed technique utilizes infrared video information and avoids imposing geometric or positional constraints on the patient. The technique also deals with fully or partially obscured patients’ body. A continuously updated breathing activity template is built for distinguishing general body movement from breathing behavior. Index Terms – breath monitoring, behavior recognition, vision analysis, respiration monitoring I. INTRODUCTION Obstructive Sleep Apnoea Syndrome is increasingly recognised as a common and serious condition, leading to sleep disturbance and consequential daytime sleepiness. Apnoea refers to short spells when breathing stops. In obstructive sleep apnoea, the throat constricts during sleep, preventing breathing; the apnoea episode often ends with a loud snore and/or gasp. The event is sufficient to open the throat muscles to allow breathing, and the patient usually falls asleep again so quickly that the event is not remembered. In obstructive sleep apnoea, this cycle repeats itself throughout the night as the muscles relax and the throat blocks again. The frequency of the episode determines the severity of the syndrome [12]. Hence, a further investigation on monitoring human breathing activity is needed for diagnosis of sleep disorders. Current breathing monitoring techniques can be c