Abstract
BACKGROUND: Accurate atrial fibrillation (AF) detection and burden assessment are critical features of modern insertable cardiac monitors (ICMs), enabling precise determination of AF episode patterns, frequency, duration, and total burden to guide treatments. OBJECTIVE: This study aimed to evaluate the AF detection performance of the Assert-IQ ICM and assess the impact of an artificial intelligence (AI) algorithm designed for reducing false-positive AF episodes. METHODS: This prospective, single-arm, multicenter study enrolled 151 subjects with symptomatic, drug-refractory paroxysmal or persistent AF. A Holter assessment was conducted after ICM insertion. AF detection metrics-sensitivity, specificity, positive predictive value (PPV), and negative predictive value-were evaluated by comparing ICM detections with core laboratory-annotated Holter AF events. The impact of an AI algorithm on AF detection performance was then assessed. RESULTS: Among 135 analyzable patients, 39 had Holter-confirmed AF with 522 episodes lasting ≥2 minutes. Assert-IQ ICM correctly identified all patients with true AF. Duration-based sensitivity, specificity, PPV, negative predictive value, and accuracy were 93.0%, 99.3%, 97.4%, 98.0%, and 97.9%, respectively. Episode detection sensitivity was 99.4% (gross) and 99.9% (patient average). AF burden correlation between ICM and Holter was excellent (r = 0.99). The AI algorithm retained all true positives and reduced 72.6% of false positives, improving PPV from 79.9% to 93.6%. CONCLUSION: Assert-IQ ICM accurately detects AF and quantifies burden for long-term monitoring. The AI algorithm effectively reduces false positives while maintaining high sensitivity.