Optimized Algorithms for Atrial Fibrillation Detection by Wearable Tele-Holter Devices

In this work we implemented and tested on a prototype tele-monitoring device an algorithmic framework for AF detection. The framework was an adaptation for device implementation of a previously developed QRS complex detector and two AF detectors based on RR irregularity features. Validation on MIT-BIH databases demonstrated the capability of the framework to provide accurate onboard AF detection with overall accuracy of 98% and affordable computation burden. The developed Tele-Holter system may have significant impact on AF prevention and screening programs as well as on the improvement of patient follow-up after cardiac events or interventions.