Cardiac arrhythmia classification using machine learning

Abstract

Machine learning models interpreting ECG signals enable scalability to increase the number of patients examined for cardiac arrhythmia and hence contribute to democratize healthcare. In this thesis a machine learning model for detecting ectopic heartbeat arrhythmias is developed and evaluated. The initial aim was to interpret signals from mobile ECG devices but due to lack of data, the model was developed and evaluated on the publicly available MIT-BIH Arrhythmia Database. The evaluation followed an inter-patient evaluation scheme to mimic the performance of the model in real-world clinical settings. The proposed model design included ensemble learning between a convolutional neural network and a multilayer perceptron network. The model was able to detect non-ectopic, supraventricular ectopic, and ventricular ectopic heartbeats with a recall of 97 %, 62 %, and 78 %, respectively.