Faculty Advisor

Chon, Ki H.

Abstract

The aim of this project was to develop a method for detecting atrial fibrillation (AF) and atrial flutter (AFL) that could be applied to clinical practice as a diagnostic test. This was achieved by using the RR-intervals of ECG data. Our AF algorithm was based on RR-interval irregularity for detection through statistical techniques: root mean squares of successive differences, turning points ratio, and Shannon entropy. For AFL detection, we implemented a time-frequency analysis of the patient data. Our AF program achieved a sensitivity of 94.9% specificity of 90.7% and accuracy of 92.4% for AF data. Our AFL program was also able to accurately determine segments of flutter with the exception of the patients with RR-intervals resembling normal sinus rhythm.

Publisher

Worcester Polytechnic Institute

Date Accepted

April 2011

Major

Biomedical Engineering

Project Type

Major Qualifying Project

Accessibility

Unrestricted

Advisor Department

Biomedical Engineering

Share

COinS