The heart is the center of the circulatory system for humans. In the heart, there are diseases or conditions that cause abnormal heart rhythms known as arrhythmias. Premature Ventricular Contraction (PVC) is an example of an arrhythmia that occurs when the ventricles have an unnaturally additional heart rate. If PVC occurs regularly, it can cause several other diseases including heart failure, coronary heart disease, and others, so it is necessary to check the health condition of the heart, namely by using an Electrocardiogram (ECG). In addition, the price of ECG medical equipment which is quite expensive and difficult to reach for some people is also the basis of this research. The author will use the triangular geometry feature and calculate the amplitude R value. The classification method applied is Artificial Neural Network (ANN) using the backpropagation algorithm. The test results of the ECG AD8232 sensor acquisition get an error value of 4.47% with 10 tests. This test compares the Beat Per Minute (BPM) value on the AD8232 and BPM sensors in real conditions. The results of the accuracy using the JST 3 hidden layer classification get 90% of the 20 test data used. For testing the computation time the system gets an average value of 235.5 milliseconds from the 20 tested data.
Copyrights © 2021