Clustering of Paroxysmal Atrial Fibrillation (PAF) and non-PAF subjects based on arrhythmia-free records


Aligholipour O., KUNTALP M.

4th Electric Electronics, Computer Science, Biomedical Engineerings' Meeting, EBBT 2018, İstanbul, Turkey, 18 - 19 April 2018, pp.1-5 identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/ebbt.2018.8391425
  • City: İstanbul
  • Country: Turkey
  • Page Numbers: pp.1-5
  • Keywords: Data clustering, ECG analysing, Fuzzy C-means, K-means, Paroxysmal Atrial Fibrillation
  • Dokuz Eylül University Affiliated: Yes

Abstract

© 2018 IEEE.Atrial fibrillation (AF) is one of the most common diseases where the atria cannot completely push the blood to ventricle and therefore clot formation could occur which may contribute to serious health problems. Paroxysmal AF is a type of AF in which the episodes could last from minutes to days and ends by itself. If PAF condition continues, it will be converted to Persistence AF. Since PAF episodes could last for a short time period, it is very difficult to record the ECG of the subjects during the arrhythmic event. The aim of this study is to analyze the structure of the data obtained from arrhythmia-free ECG records of PAF and non-PAF subjects by clustering. In other words, the separability of the two types of data is to be investigated. The obtained results show a significant overlap between the two types and consequently a good classifier is needed.