Deteksi Myocardial Infarction Menggunakan Fitur Statistik Segmen-ST Elektrokardiogram dan Analisis Diskriminan

This article describes the detection of myocardial infarction using the statistical mean features, including the mean, median, and standard deviation. The classification used is the discriminant analysis method which is implemented using matlab software. The ECG signal obtained from the device is then processed. After that, the feature extraction is carried out. The results of the extraction is normalized so that all patient data have the same standard in amplitude wave magnitude. After normalization, the data will be used as input for discriminant analysis. In this article we try to use the mean, median, and standard deviation features. In this experiment using 15 leads consisting of 12 conventional leads and 3 posterior leads, the addition of these 3 leads has the advantage of determining the performance results obtained. Percentage of accuracy performance, the best percentage of accuracy performance is 97.73% with the mean feature. This experiment tries to compare the features of mean and standard deviation, mean and median, standard deviation and median, and mean, median, and standard deviation. The combined experiment shows that the best accuracy performance percentage value is 98.84% with standard deviation and median features.