Automatic sleep apnea detection and sleep classification using the ECG and the SpO2 signals

Dissertation for a Masters Degree in Computer and Electronic Engineering

The present work describes the aspects to implement a system that can be used as a swift and accessible screening tool in patients whose complaints are compatible with OSAS (Obstructive Sleep Apnea Syndrome). This system only uses two signals, electrocardiogram (ECG) and the saturation of oxygen in arterial blood flow (SPO2). This system would be applied for the ambulatory automatic screening of OSAS, which currently are done in a Hospital environment, with a substantial waiting list. The system also would overcome the time consuming visual sleep scoring that contributes for the mentioned waiting list. We have developed a system that automatically detects OSAS based on the ECG and SpO2. However this system has to be paired up with another that detects the awake/sleep/REM periods (also based on the ECG), which is also part of this work. This last component has proved to produce results that are complex to classify,for which there is still a lack of research work. However we have described the necessary algorithms, and have used state-of-the-art signal processing tools, such as wavelets.