Detection of sleep disordered breathing severity ((Obstructive sleep apnea-hypopnea syndrome) using acoustic biomarker and machine learning techniques
Detection of sleep disordered breathing severity ((Obstructive sleep apnea-hypopnea syndrome) using acoustic biomarker and machine learning techniques
Wahab Aminiazar1 Rasoul Farahi2
1) 1 Assistant Professor, Department of Electrical and Computer Engineering, Islamic Azad University, Mahabad Branch
2) Frahi Instructor, Department of Computer Engineering, Islamic Azad University, Mahabad Branch, Mahabad, Iran
IOI Code
:
XDHZ-FCBAH
Publication :
3 rd Conference on Mechanical, Electrical and Computer Engineering (mecconf.com)
Abstract :
Obstructive sleep apnea-hypopnea syndrome is one of the known sleep disorders that causes daytime drowsiness and is associated with diseases such as type 2 diabetes, cardiovascular disease and stroke. Polysomnography testing is an old method for diagnosing obstructive sleep apnea-hypopnea syndrome. While this test is expensive, it is inconvenient and requires the placement of body contact sensors during sleep. Recently, several studies have used snoring and bloating obtained by non-contact microphones to diagnose obstructive sleep apnea-hypopnea syndrome. Obstructive sleep apnea-hypopnea syndrome may be slightly audible to loud. Diagnosis of obstructive sleep apnea-hypopnea syndrome, especially low-intensity obstruction-hypopnea syndrome due to low signal-to-noise ratio, can be challenging in noisy environments. In this article, we propose a new method for the rapid diagnosis of obstructive sleep apnea-hypopnea syndrome (sleep-disordered breathing) from sleep-recorded data. The proposed method is based on the artificial neural network method in which the post-diffusion learning algorithm is used. The proposed method can diagnose obstructive sleep apnea-hypopnea syndrome with 98.178% accuracy compared to previous methods.
Keywords : obstructive sleep apnea hypopnea syndrome respiratory disorders acoustic biomarkers machine learning techniques