Abstract
dc:description.abstractPolysomnography, plays a crucial role in diagnosing a wide range of sleep-related conditions, including obstructive sleep apnea and central sleep apnea. However, traditional polysomnography requires lab setups with multiple sensors in an overnight study, which makes it resource-intensive and inconvenient for both patients and healthcare providers. In this work, we investigated whether combining pressure-sensitive mat data with other minimally invasive sensors and deep learning techniques can provide accurate and reliable analyses comparable to traditional polysomnography setups. We finetuned Vision Transformers for accurate sleep position estimation using pressure-sensitive mat data, illustrating their adaptability from natural images to clinical applications. We developed a multimodal deep learning approach with ImageBind, combining pressure-sensitive mat data and electrooculography signals to improve sleep stage classification by capturing complementary information across modalities. Lastly, we demonstrated that by fusing knowledge of sleep position with respiratory effort data, we can improve the detection of sleep apnea and hypopnea.
Degree
thesis:*- Name thesis:degree_name
- Master of Applied Science (M.App.Sc.)
- Level thesis:degree_level
- Master's
- Discipline thesis:degree_discipline
- Engineering, Electrical and Computer
- Grantor dc:publisher
- Carleton University
- Year dc:date.issued
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Papillon, Olivier Lucas
Subjects
dc:subject × 3Rights
dc:rights- Statement dc:rights
-
- Copyright © 2025 the author(s). Theses may be used for non-commercial research, educational, or related academic purposes only. Such uses include personal study, distribution to students, research and scholarship. Theses may only be shared by linking to the Carleton University Institutional Repository and no part may be copied without proper attribution to the author; no part may be used for commercial purposes directly or indirectly via a for-profit platform; no adaptation or derivative works are permitted without consent from the copyright owner.
- Language dc:language.iso
- en