<![CDATA[91% of sudden deaths in Indonesia are due to Arrhythmias. Atrial Fibrillation is one type of arrhythmia that occurs most often among other types. Atrial fibrillation is characterized by an irregular heartbeat due to abnormal electrical activity. Blood collects in the atria and is not supplied to the ventricles sufficiently. So that the heart fails to pump enough blood to the lungs. For the examination, it is quite expensive, especially in the hospital. Therefore a system was developed to be able to detect AF without injuring the body. The system consists of several functions, namely the ECG signal generator using the AD8232 sensor, data processing using the Arduino Nano board, and displaying the classification or diagnosis results of the class "AF" or "Normal" using an LCD. In the classification process, statistical features in the form of Mean, Median, Standard Deviation, Min and Max are used as test data from the K-Nearest Neighbor Method. This method is used because it does not require a training process. Each test was carried out as much as 20 times. The test results are in the form of 96.83% for the sensor accuracy level. Accuracy of 95%, 90%, and 85% for the accuracy of K-NN classification based on k = 3, 5, and 7 so that the best accuracy is achieved with k = 3.In addition, there is also a relatively short computation time of 16.44 ms so that users it does not take long to find out the results of the classification or the results of the diagnosis.]]>
Copyrights © 2021
