Diagnosing Heart Disease using Wavelet Transformation and Adaptive Neuro Fuzzy Inference System (ANFIS) Based on Electrocardiagram (ECG)

Abstract

Heart disease is the leading cause of death in the world. Heart disease is called the silent killer, because it often occcurs suddenly. Therefore, periodic cardiac examination is very necessary to reduce cases of death from heart disease.Heart disease can be known through electrocardiogram (ECG) examination. This study aims to explain the process of diagnosing heart disease through ECG using wavelet transformation and Adaptive Neuro Fuzzy Inference System (ANFIS).

The process of diagnosing heart disease begins with cutting ECG signal consisting of 9-11 waves into one ECG wave, then decomposition and extraction are performed using wavelet transformation to obtain 6 parameters. The parameters will be used as input in ANFIS model. Data obtained from ECG extraction are divided into 70% training data and 30% testing data The output from the ANFIS model is a diagnosis of heart diseases, such as left bundle branch block (LBBB),  right bundle branch block (RBBB), and normal. ANFIS learning is divided into 6 stages, namely clustering data with Fuzzy C-Means method, computing the degree of membership of each data, determining fixed neurons, looking for normalized firing strength, calculating the consequent parameter values, and determining network output.

The results of the study obtained the best ANFIS model with 10 clusters. The level of accuracy, specificity, and sensitivity for training data is 100%, 100%, and 100%, respectively and for the testing data, the level of accuracy, specificity, and sensitivity is 100%, 100%, and 100%, respectively.