Identification of atrial fibrillation using descriptive statistic features and adaptive Neuro-Fuzzy inference system

Abstract

Atrial fibrillation (AF) is one of the most common arrhythmia which can cause a serious problem. Nevertheless, well treated AF might not lead to any further complication. Early detection of AF could be an important preventative step that have to be conducted. In this article, we aim to make an automatic detection of atrial fibrillation. Seven descriptive statistic features have been utilized to detect AF. The features obtained could differ between two condition: normal and AF. Later, we use Adaptive Neuro-Fuzzy Inference System (ANFIS) as a classification method. Sugeno-type fuzzy inference system along with Gaussian type of fuzzy are utilized to classify the condition. The proposed method is applied on MIT-BIH Atrial Fibrillation database. The performance obtained from this proposed method might be considered for a medical application.