Atrial fibrillation (AF) is the most common cardiac arrhythmia and a leading cause of ischemic stroke. Several healthcare services relying on remote ECG monitoring have been proposed for real-time detection and diagnosis of silent AF. The goal of this project is to lay down the clinical and technical basis of a wearable ECG recorder provided with novel artificial intelligence-assisted algorithms for the diagnosis of atrial fibrillation.
Primary goals:
1. publication of the results on the performance of the algorithm on public ECG libraries;
2. algorithm validation on a setting of real life Holter-derived ECG traces;
3. development of a prototype of a SAFE MATE ECG device for the diagnosis of AF in the operating environment.

The group has developed a novel multi-lead sub-beat ECG-based algorithm for atrial fibrillation detection using machine learning and is finalizing the related edge computing enabled prototype.

Atrial Fibrillation Detection by Multi-Lead ECG Processing at the Edge A. Petroni, F. Cuomo, G. Scarano, P. Francia, S. Colonnese
2021 IEEE Globecom Workshops (GC Wkshps), 1-6