Detection of Myocardial Ischaemia by using ECG, Artificial Neural Network and Gaussian Mixture Model

In this study, a new technique which detects anomalies in skin sympathetic nerve activity (SKNA) and ECG by using state-of- the-art signal processing and machine learning methods is developed to perform robust detection of myocardial ischaemia (MI). For this purpose, a signal processing technique that simultaneously obtains SKNA and ECG from wideband recordings on STAFF III database is developed. By using preprocessed data, a novel feature extraction technique which obtains ECG features that are critical for reliable detection of MI is developed. By using extracted features, a supervised learning technique based on artificial neural network (ANN) and an unsupervised learning technique based on Gaussian mixture model (GMM) are developed to perform robust detection of ECG anomalies. A Neyman-Pearson type of approach is developed to perform robust detection of outliers that correspond to MI. The performance results of proposed technique over STAFF III database showed that technique provides highly reliable detection of MI by performing robust detection of ECG anomalies. Therefore, in cases where diagnostic information of ECG is not sufficient for reliable diagnosis of MI, proposed technique can be used to provide early and accurate diagnosis of the disease, which can lead to a significant reduction in mortality rates of ischaemic heart diseases.