<![CDATA[Atrial fibrillation is a type of arrhythmia that occurs due to electrical breakdown from the atria so that blood cannot be pumped to the ventricles at the right time. This condition can lead to other complications, such as stroke, palpitations, cardiomyopathy and heart failure. The detection can be done using an electrocardiograph (ECG), holter monitor or electrophysiology, but this action takes a lot of time and money. Based on this, research was carried out to detect Atrial Fibrillation as soon as possible by building a system consisting of sensor AD8232, Arduino Uno and 16x2 LCD. The ECG signal will be acquired by the AD8232 sensor and processed by Arduino Uno to obtain feature values. The features used are the average and the median value of the QRS gradient and interval. The Artificial Neural Network (ANN) method is used as a classifier of these features with 2 conditions, “Normal†and “FAâ€. A total of 60 datasets were used, 40 of which were used as training data in the ANN training phase with backpropagation algorithm and 20 were used as testing data. From the BPM test carried out by comparing the sensor acquisition value and the manual calculation value, it was found that the accuracy of the sensor in acquiring the ECG signal was 94,55%. Then from the 20 tested data, the classification accuracy of the ANN method is 90% with an average calculation time of 32,09 ms.]]>
Copyrights © 2021
