Open Collections

Grooby, Ethan

University of British Columbia

There is an urgent need to make health monitoring, particularly cardio-respiratory monitoring, easily accessible to parents and newborns not covered by professional medical care and equipment in the critical first month of life. Hence, the overall aim of this thesis is the development of supporting software for an affordable, non-invasive solution for monitoring neonatal cardio-respiratory health automatically. This is achieved through audio analysis of digital stethoscope-recorded chest sounds and video analysis of camera-recorded neonates. For audio analysis, chest sounds are processed through state-of-the-art methods of automated signal quality analysis and sound separation to obtain high-quality heart and lung sounds. Heart and breathing rate estimation methods were then developed. Utilising these methods, automated health assessment was explored in the context of (1) clinical outcome identification, (2) murmur detection, (3) respiratory distress prediction, and (4) lung aeration assessment. For video analysis, extraction of photoplethysmogram and heart rate were demonstrated in clinically obtained neonatal videos. This was achieved through the development of a state-of-the-art face detector, in combination with an adapted region of interest tracker. Skin regions were then identified from the tracked face, from which photoplethysmogram and subsequently heart rate were extracted. Overall this thesis proposes state-of-the-art methods in the fields of neonatal audio and video analysis applied to cardio-respiratory monitoring.

Text

2024-01-12

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/87233

Electrical and Computer Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/