Sleep Apnea Screening

Grade Level at Time of Presentation

Senior

Major

Computer Science

Minor

Not declared

Institution

Morehead State University

KY House District #

73

KY Senate District #

28

Department

Computer Science

Abstract

Sleep Apnea Screening Abstract

Authors: Andrew S. Cooper, Cody J. Mitchell

Mentor: Shahrokh N. Sani

Department: Computer Science

Recent studies reveal there is a strong correlation between sleep apnea and hypertension and cardiovascular disease. It is well known that Obstructive Sleep Apnea (OSA) increases the risk of atrial fibrillation and relates to congestive heart failure, and other vascular diseases. Traditionally, sleep apnea has been diagnosed via overnight polysomnography (PSG). PSG is inconvenient, expensive and limited access test, therefor, in many cases sleep apnea passes undiagnosed. The yearly financial cost of serious OSA in the United States is 65 to 165 billion dollars. Because overnight PSG is inconvenient, expensive, and limited clinical based service. There is a large interest in developing alternative methods of identifying OSA. In this regard, we have developed a simple system that provides an easy, reliable, inexpensive, and transportable approach to automate the diagnosis of sleep apnea. This allows the public to have OSA tests at home using their smart phone without the need for attended overnight tests. The approach uses smartphone wearable technology (as a data acquisition platform for collecting electrocardiogram (ECG) signal) and machine learning classification as an analysis technique. We investigated new features of ECG signals which is alternated with Obstructive Sleep Apnea. The new cardiovascular variable was used as a new attribute in our comprehensive Obstructive Sleep Apnea prediction model to increase its accuracy to predict the class label of unknown patient records (sleep apnea vs non-sleep apnea).

This document is currently not available here.

Share

COinS
 

Sleep Apnea Screening

Sleep Apnea Screening Abstract

Authors: Andrew S. Cooper, Cody J. Mitchell

Mentor: Shahrokh N. Sani

Department: Computer Science

Recent studies reveal there is a strong correlation between sleep apnea and hypertension and cardiovascular disease. It is well known that Obstructive Sleep Apnea (OSA) increases the risk of atrial fibrillation and relates to congestive heart failure, and other vascular diseases. Traditionally, sleep apnea has been diagnosed via overnight polysomnography (PSG). PSG is inconvenient, expensive and limited access test, therefor, in many cases sleep apnea passes undiagnosed. The yearly financial cost of serious OSA in the United States is 65 to 165 billion dollars. Because overnight PSG is inconvenient, expensive, and limited clinical based service. There is a large interest in developing alternative methods of identifying OSA. In this regard, we have developed a simple system that provides an easy, reliable, inexpensive, and transportable approach to automate the diagnosis of sleep apnea. This allows the public to have OSA tests at home using their smart phone without the need for attended overnight tests. The approach uses smartphone wearable technology (as a data acquisition platform for collecting electrocardiogram (ECG) signal) and machine learning classification as an analysis technique. We investigated new features of ECG signals which is alternated with Obstructive Sleep Apnea. The new cardiovascular variable was used as a new attribute in our comprehensive Obstructive Sleep Apnea prediction model to increase its accuracy to predict the class label of unknown patient records (sleep apnea vs non-sleep apnea).