A Neuro-Fuzzy Identification of ECG Beats

Abstract

This paper presents a fuzzy rule based classifier and its application to discriminate premature ventricular contraction (PVC) beats from normals. An Adaptive Neuro-Fuzzy Inference System (ANFIS) is applied to discover the fuzzy rules in order to determine the correct class of a given input beat. The main goal of our approach is to create an interpretable classifier that also provides an acceptable accuracy. The performance of the classifier is tested on MIT-BIH (Massachusetts Institute of Technology-Beth Israel Hospital) arrhythmia database. On the test set, we achieved an overall sensitivity and specificity of 97.92% and of 94.52% respectively. Experimental results show that the proposed approach is simple and effective in improving the interpretability of the fuzzy classifier while preserving the model performances at a satisfactory level.

Ectopic beats detection and correction methods: A review

“…Chikh et al [54] developed an adaptive network-based fuzzy inference system (ANFIS) in order to classify the PVC beats. Authors aimed to increase the interpretability and understandability of the classification by using ANFIS classification approach.…”

Novel Beat-wise Classification Algorithm of ECG to Monitor Progression of Cardiac Health

“…The performance of our classifier for detecting normal beats is given in Table VII. The model-based fuzzy classifier was compared to other ECG beat classifiers such as neuro-fuzzy [17], fuzzy logic system [16], deep learning [32], and wavelet-Support Vector Machine (SVM) [33] in Table VIII. BIDMC CHF database is a Class III database available on Physionet that contains long-term ECG from patients with severe CHF, NYHA class 3-4.…”

“…Sensors that monitor potassium blood content, activity-related energy consumption, skin temperature, and sweating can alert the consumer during HF [15]. ECG beat classification can be performed using various techniques such as fuzzy logic [16], neuro-fuzzy [17], adaptive fuzzy neuro systems using Lyapunov exponents [18], and extreme learning machine algorithms [19]. The output block consists of plots or medical questionaries' regarding heart attack symptoms [20].…”

Localization of premature ventricular contraction foci in normal individuals based on multichannel electrocardiogram signals processing

“…PVC is an extra heart beat originates from the ventricles and comes before the normal heart beat. Although in general, this arrhythmia may occur in a healthy person, but it is mostly associated with elderly patients (males: 60-80%) (Nathani et al 2007) and patients suffering cardiac diseases such as hypertension, myocardial infarction, and so on (Chikh et al 2010). …”