Deteksi Fibrilasi Atrium Menggunakan FAM yang Dikombinasikan dengan Grid Search

Atrial Fibrillation (AF) is most found arrhythmias in daily medical practice. AF has correlation with other medical conditions and gives risk of stroke to the patient. Research for AF is interesting for multi discipline of science. One of them is computational sciences with the purpose to develop an AF detection. There are three characteristics of AF appearance. The irregularity of RR-intervals, the loosing of P wave, and the appearing of fibrillations wave. Among that three, the irregularity of RR intervals is the most used as the key feature. In this article, the irregularity of RR interval is detected using Finding Anomalies Around the Mean (FAM). FAM is the method to detect irregularity by comparison of data based on the mean of data in the observed set. FAM has x and y parameters that need to be fixed. For that purpose, FAM is combining by grid search. This combination shows us a good performance, because the system of grid search works thoroughly and regularly. This method has been used to test the 22 patient’s data from MIT-BIH. Using the approach and the clinical data, accuracy, sensitivity, and specificity are obtained, whice are 92.63%, 95.38%, and 90.57% respectively.