Automated Real-Time Atrial Fibrillation Detection on a Wearable Wireless Sensor Platform

This paper presents an automated real-time atrial fibrillation (AF) detection approach that relies on the observation of two characteristic irregularities of AF episodes in the electrocardiogram (ECG) signal. The results generated after the analysis of these irregularities are subsequently analyzed in real-time using a new fuzzy classifier. We have optimized this novel AF classification framework to require very limited processing, memory storage and energy resources, which makes it able to operate in real-time on a wearable wireless sensor platform. Moreover, our experimental results indicate that the proposed on-line approach shows a similar accuracy to stateof- the-art off-line AF detectors, achieving up to 96% sensitivity and 93% specificity. Finally, we present a detailed energy study of each component of the target wearable wireless sensor platform, while executing the automated AF detection approach in a real operating scenario, in order to evaluate the lifetime of the overall system. This study indicates that the lifetime of the platform is increased by using the proposed method to detect AF in real-time and diagnose the patient with respect to a streaming application that sends the raw signal to a central coordinator (e.g., smartphone or laptop) for its ulterior processing.


Published in:
Proceedings of 34th IEEE Annual International Conference of the Engineering in Medicine and Biology Society (EMBC 2012), 1, 1, 2472-2475
Presented at:
34th IEEE Annual International Conference of the Engineering in Medicine and Biology Society (EMBC 2012), San Diego, USA,, August 28-September 1, 2012
Year:
2012
Publisher:
New York, IEEE Press
ISBN:
978-1-4244-4119-8
Keywords:
Laboratories:




 Record created 2012-06-04, last modified 2018-09-13

Publisher's version:
Download fulltext
PDF

Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)