E-Locus - Institutional Repository of the University of Crete

Home Collections School/Department School of Medicine Department of Medicine Post-graduate theses

Post-graduate theses

Current Record: 26 of 1210

[Add to Basket]

Identifier

000457827

Title

Automatic obstructive sleep apnea detection

Alternative Title

Αυτόματος εντοπισμός υπνικής άπνοιας

Author

Μπάντρα, Κωνσταντίνα

Thesis advisor

Σάκκαλης, Βαγγέλης

Abstract

Over the years, ballistocardiography (BCG) has emerged as a unob- structive, non-invasive, and safe technique that can be used for cal- culating the heart rate of a patient. As the heart rate is an essential element not only for the differential diagnosis process but also for mon- itoring patients, various scholars have conducted research on possible applications and use for the BCG signal. Among these, sleep apnea has gained a lot of interest as it benefits greatly from the unobstructive nature of the BCG recording. Compared to Polysomnography (PSG), the golden standard used today, BCG bears the advantage of record- ing the heart rate and the respiration rate without physically touching the patient, thus removing the discomfort of the multiple attached wires and sensors which are needed during the PSG. This is crucial, as for an accurate diagnosis, multiple hours of recordings have to be acquired and most of the time the procedure has to be repeated for more than one night. By removing the discomfort for the patient we could obtain more accurate results and people would be less hesitant to undergo the diagnosis procedure. Furthermore, as the BCG sensor is a very small and portable device, at home monitoring could also be a possibility. With the present thesis we aim to assess the possibility of monitoring the heart rate and respiration rate of a patent using solely a BCG recording. Additionally, the possibility of developing a system which can utilize the calculated physiological signals and automat- ically detect Obstructive Sleep Apnea Events during patient’s sleep, using machine learning and Artificial Intelligence techniques is also explored hoping that the system will, in the long run, help in reducing diagnostic ambiguity and also in creating more concise and realizable results

Language

English

Issue date

2023-07-28

Collection

School/Department--School of Medicine--Department of Medicine--Post-graduate theses

Type of Work--Post-graduate theses

Permanent Link

https://elocus.lib.uoc.gr//dlib/1/1/e/metadata-dlib-1693646477-437516-11328.tkl

Views

414

Digital Documents
	Download document View document Views : 2