A Novel System for the Rapid and Automated Detection of Atrial Fibrillation

Aim: Telemetry monitoring with accurate automated rhythm classification can improve diagnosis and management of patients with atrial fibrillation (AF). The objective of this study is to perform the initial clinical validation of a new e-Health device and algorithm developed to distinguish automatically, fast and accurately AF episodes from sinus rhythm (SR).

Methods and Results: A hand-held electrocardiogram recording device (Maestro) and a signal processing platform were developed. Signals were acquired from 66 consecutive patients along with a standard ECG. Six-second of Maestro recordings and ECG signals from the 66 patients were analyzed independently using the Maestro algorithm and by 3 expert physicians. The Maestro calculated dimensionless temporal R-R interval variability (VRR) index, a spectral frequency dispersion metric (FDM), and a Gaussian Mixed Model (GMM) that were developed and utilized to accurately diagnose AF. The 2-dimensional scatter-gram of the Maestro metrics demonstrated 2 distinct clusters of VRR and FDM in patients with SR or AF. The VRR index clusters in patients with SR and AF were 0.018±0.013 and 0.187±0.073 (mean±SD), respectively (P<0.001). The FDM clusters in patients with SR and AF were observed at 10.5±5.916 and 15.892±3.337, respectively (P<0.001). The GMM algorithm correctly categorized the AF (N=46) and SR (N=20) for all Maestro segments analyzed with 100% specificity and sensitivity as compared to the standard ECG interpreted by the blinded electrophysiologists.

Conclusions: The Maestro handheld device utilizes a novel classification algorithm and was demonstrated to acquire and automatically analyze short 6-second electrograms for rapid and accurate diagnosis of AF.