
Background: Noninvasive monitoring of atrial fibrillation (AF) remains a significant clinical challenge. Furthermore, differentiating AF from related atrial tachyarrhythmias (AFL/AT) is essential for guiding therapy. Devices based on photoplethysmography (PPG) are limited by their bulky design and inadequate accuracy, diminishing their clinical applicability. This study aimed to evaluate a novel, wearable, PPG-based device and proprietary algorithm for detecting and classifying AF and related arrhythmias.
Methods: This prospective study included patients undergoing catheter ablation for AF or AFL/AT. A miniature PPG sensor (Cardiodet LTD, Israel) was applied to the left fifth digit. Waveforms were transmitted wirelessly, analyzed offline using a proprietary algorithm, and validated against 12-lead ECG and coronary sinus catheter recordings. Sensitivity and specificity for atrial arrhythmia detection, as well as AF differentiation from AT/AFL, were calculated.
Results: Sixteen patients undergoing catheter ablation for AF were enrolled, resulting in an overall recording time of 1498 minutes. 298 arrhythmia episodes (225.4 minutes; 15% arrhythmia burden) were documented. Based on the PPG waveform, the algorithm detected 271 episodes (91%) as atrial arrhythmias, with sensitivity and specificity of 0.91 and 0.95, respectively (ROC AUC 0.971, Figure 2). The device reliably differentiated AF from AT/AFL, with a sensitivity of 0.91 and specificity of 0.93.
Conclusion: In patients undergoing catheter ablation for AF and related arrhythmias, a novel PPG-based miniature device demonstrated high accuracy in detecting AF and differentiating it from AFL/AT. Its compact design and robust algorithm facilitate enhanced atrial arrhythmia monitoring and may improve patient compliance.

