Ventricular arrhythmia detection with a wearable ring-type photoplethysmography sensor: A feasibility study

BACKGROUND: Ventricular arrhythmias (VAs) are most frequently the onset of sudden cardiac death, and early detection is key for survival. Limited studies suggest that wrist-worn photoplethysmography (PPG) sensors can detect VA. Ring-type wearable PPG sensors may be more accurate; however, they have not yet been investigated. OBJECTIVE: We aimed to show the sensitivity of a ring-type wearable PPG sensor to detect induced VA. METHODS: Patients undergoing a procedure involving VA induction were included. The CART-I was positioned on the finger during the procedure, and PPG and reference electrocardiogram data were collected. A loss of cardiac output was defined as a flatline, or a decrease in amplitude of ≥50%. Sensitivity was calculated using a contingency table. Bland-Altman analysis and intraclass correlation coefficients were used to assess the agreement between PPG- and electrocardiogram-derived arrhythmia durations. RESULTS: Thirty-seven patients were enrolled, of whom 25 were included: defibrillation threshold testing (n = 16), ventricular tachycardia (VT) ablation (n = 8), and electrophysiological study (n = 1). A total of 67 VA episodes, of which 27 were ventricular fibrillation (VF) and 40 were VT episodes, were recorded. A loss of cardiac output was observed on the tachograms of all VF episodes and 36 VT episodes (90%). The sensitivity was 94% (VF, 100%; VT, 90%). The mean difference in VA duration was 0.1 seconds (95% limits of agreement -1.4 to 1.7), with an intraclass correlation coefficient of 0.998 (95% confidence interval, 0.997-0.999), indicating excellent agreement. CONCLUSION: This study suggests that a ring-type wearable device can detect, with high sensitivity, changes in arterial pulse flow during induced VA.