BME Seminar: Jiayan Huo
Monday, April 4th, 2022, 12:00 p.m.
Jiayan Huo
BME Ph.D. Candidate
Dr. Janet Roveda Lab
“Home-Based Sleep Evaluation and Sleep-Disordered Breathing Screening Using Machine Learning”
Keating 103
Live Zoom | Passcode: BearDown
(Instructor permission required for enrolled students to attend via Zoom)
Hosts: Dr. Beth Hutchinson and Dr. Russ Witte
Persons with a disability may request a reasonable accommodation by contacting the Disability Resource Center at 621-3268 (V/TTY).
ABSTRACT:
About 50 to 70 million Americans suffer from insufficient sleep and irregular sleep-wake behaviors. Chronic sleep deprivation can lead to other comorbidities burdens, such as cardiovascular diseases, hypertension, and diabetes. Due to the large undiagnosed sleep disorders population in the U.S., easy-to-use screening tools and proper home testing are required. As a powerful set of tools for data analysis, machine learning can automatically recognize patterns and predict unseen data and has shown promise in biomedical data analysis. This talk introduces the machine learning technique to home-based obstructive sleep apnea (OSA) screening and sleep evaluation.
OSA is one of the most common sleep-disordered breathing and has a significant negative impact on health. A machine learning-based questionnaire was developed to classify OSA risk. The questionnaire that incorporates OSA subtype information improves the accuracy of OSA screening compared to other commonly used screening instruments. It has the potential to be an essential clinical tool in the identification of patients with OSA.
Further, a multi-task deep learning algorithm was developed for home-based sleep evaluation using single-lead electrocardiography (ECG). Specifically, the model can extract features from the raw ECG signal, detect cortical arousals at the one-second resolution, and score sleep into four stages every 30 seconds. It is inconvenient to record an electroencephalogram (EEG) at home for arousal detection. The algorithm provides the potential for home-based type III sensors to identify cortical arousals through ECG. It can assist analysis of the correlation between arousal and autonomic nervous system activity by combining the detected arousal and scored sleep stages.