Automated detection of atrial fibrillation (AF) is an interesting topic. It is an account of very dangerous, not only as a trigger of embolic stroke, but it’s also related to some else chronical disease. In this study, we analyse the presence of AF by determining irregularities of RR-interval. We utilize the interval comparison to measure the degree of irregularities of RR-interval in a defined segment. The series of RR-interval is segmented with the length of 10 of them. In this study, we use interval comparison for the method. We were comparing all of the intervals there each other. Then we put the threshold to define the low difference and high difference (δ). A segment is defined as AF or Normal Sinus by the number of high δ, so we put the tolerance (β) of high δ there. We have used this method to test the 23 patients data from MIT-BIH. Using the approach and the clinical data we find accuracy, sensitivity, and specificity of 84.98%, 91.99%, and 77.85% respectively. © Published under licence by IOP Publishing Ltd.
Eliana M., Nuryani N.
Journal of Physics: Conference Series, 10.1088/1742-6596/909/1/012020