Atrial Fibrillation Detection Using Swarm Fuzzy Inference System and Electrocardiographic P-Wave Features

A new technique for detecting atrial fibrillation (AF) is proposed and investigated. The technique employs a swarm fuzzy inference system (SFIS). SFIS is fuzzy system optimized by using a particle swarm optimization (PSO). The technique introduces new inputs for the SFIS to detect AF. The inputs involve the peaks number and width of electrocardiographic P-wave. Experiments of FA detection utilizing SFIS with different inputs are conducted. On a test using clinical electrocardiographic data, SFIS performs well in AF detection with sensitivity, specificity and accuracy of 77.86%, 60.40% and 75.09%, respectively.