Machine Learning Approaches for Ambulatory Electrocardiography Signal Processing

The ambulatory electrocardiography (AECG) records the ECG while the patient is doing real-life activities. It allows the study of transient phenomena and cases of fatal arrhythmic events, including sudden cardiac death. However, noise and artifacts can corrupt the AECG signal which downgrades the underlying diagnostic information. This research focuses on the development of new machine-learning-based methods for improving the processing of the AECG signal. The relevance of this topic resides on the fact that improved processing steps may lead to reliable markers, thereby decreasing the risk of an incorrect diagnostic.
The first topic addressed in this book is the problem of ectopic heartbeat detection in the AECG as preprocessing step for heart rate variability or QT interval analyses. In this context, supervised learning algorithms based on support vector machines were evaluated. The new algorithms use tensors and tensor decompositions to deal directly with multi-lead AECG recordings. This approach is effective and saves training time since only one classifier is trained for each record. Furthermore, high performances were obtained considering only small training sets.

The next step covered in this work is the detection of the T-wave end in the AECG. Here, supervised learning algorithms based on neural networks and support vector machines were evaluated. Then, a novel algorithm based on support vector machines is presented for detecting the T-wave end. The new approach does not require large datasets for training and includes a robust and effective algorithm for selecting the training set. Moreover, extended evaluation and comparison of the proposed approach against state-of-the-art techniques are presented and discussed. The results showed that the proposed algorithm outperforms the state-of-the-art methods.

Finally, this research presents a software tool for the analysis of the QT interval in the AECG. The software was developed for cardiologists and specialists, and no programming skills are needed to use it. Since QT markers are related to risk stratification of suffering life-threatening arrhythmias and sudden cardiac death, this tool constitutes a useful input to QT analysis. In this context, it will support the research on ventricular repolarization analysis.