FEATURE EXTRACTION OF HEART SIGNALS USING FAST FOURIER TRANSFORM

Abstract

Heart disease is a disease that is very dangerous. Even today in Indonesia an estimated 20 million or approximately 10% of the population of the archipelago suffer from heart disease. Such conditions make cardiovascular disease the number one killer in Indonesia. Not only in Indonesia, the number of heart patients in the world very much, it is estimated there are at least one billion people. This study was designed with the aim to classify the heart signals, the data taken from Physiobank namely MIT-BIH Arrhythmia Database and MIT-BIH Normal Sinus Rhythm Database, the data is processed by taking the feature extraction methods Fast Fourier Transform. The results of the feature extraction method used will be selected prior to use for the classification process. Classification is using Backpropagation Neural Network. From the research found, the feature extraction method of Fast Fourier Transform by taking 64 point data after FFT process and backpropagation as classification of obtaining a classification accuracy rate of 87 %.