Artificial intelligence–enhanced electrocardiography and sudden death risk factors in cardiac arrest survivors diagnosed with hypertrophic cardiomyopathy

Le résumé fourni par la source

Background: The characteristics of patients with hypertrophic cardiomyopathy (HCM) who present with sudden cardiac arrest as the initial manifestation of the disease are not well known. Objective: The purpose of this study was to evaluate artificial intelligence-enhanced electrocardiography (AI-ECG) and sudden cardiac death (SCD) risk factors in patients with HCM presenting with sudden cardiac arrest (SCA). Methods: We identified patients within the Mayo Clinic enterprise (2001-2022) who were newly diagnosed with HCM after surviving SCA. Index clinical, electrocardiographic, and imaging characteristics were documented. Risk models for SCD in HCM were retrospectively applied. An AI-ECG algorithm designed for the detection of HCM was also applied to the patients' first ECG obtained at our institution. Results: Twenty-seven patients met the inclusion criteria. Eight patients (30%) had documented cardiac symptoms preceding SCA. Twenty-four patients (89%) had an abnormal electrocardiogram at index evaluation, and 16 (59%) had ≥1 SCD risk factor that could have qualified them for an implantable cardioverter-defibrillator if HCM had been diagnosed before SCA. Retrospective application of the European Society of Cardiology 5-year HCM Risk-SCD tool yielded a median score of 4.6% (interquartile range 3.2%-7.2%), with 12 patients (44%) having an estimated SCD risk of <4%, implying a low risk. AI-ECG indicated a high prediction score for HCM in 26 patients (96%), suggesting its potential utility as an early detector of the disease if applied before SCA. Conclusion: Among patients with SCA who were then newly diagnosed with HCM, conventional SCD risk factors were common but not universal in this post hoc assessment. AI-ECG may facilitate the early detection of HCM.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.