Artificial intelligence-guided detection of under-recognised cardiomyopathies on point-of-care cardiac ultrasonography: a multicentre study

Background Point-of-care ultrasonography (POCUS) enables cardiac imaging at the bedside and in communities but is limited by abbreviated protocols and variation in quality. We aimed to develop and test artificial intelligence (AI) models to screen for under-diagnosed cardiomyopathies from cardiac POCUS. Methods In a development set of 290 245 transthoracic echocardiographic videos across the Yale-New Haven Health System (YNHHS), we used augmentation approaches, and a customised loss function weighted for view quality to derive a POCUS-adapted, multi-label, video-based convolutional neural network that discriminates hypertrophic cardiomyopathy and transthyretin amyloid cardiomyopathy from controls without known disease. We evaluated the model across independent, internal, and external, retrospective cohorts of individuals undergoing cardiac POCUS across YNHHS and the Mount Sinai Health System (MSHS) emergency departments (between 2012 and 2024) to prioritise key views and validate the diagnostic and prognostic performance of single-view screening protocols. Findings Between Nov 1, 2023, and March 28, 2024, we identified 33 127 patients (mean age 58<middle dot>9 [SD 20<middle dot>5] years, 17 276 [52<middle dot>2%] were female, 14 923 [45<middle dot>0%] were male, and for 928 [2<middle dot>8%] sex was recorded as unknown) at YNHHS and 5624 patients (mean age 56<middle dot>0 [20<middle dot>5] years, 1953 [34<middle dot>7%] were female, 2470 [43<middle dot>9%] were male, and for 1201 [21<middle dot>4%] sex was recorded as unknown) at MSHS with 78 054 and 13 796 eligible cardiac POCUS videos, respectively. AI deployed to single-view POCUS videos successfully discriminated hypertrophic cardiomyopathy (eg, area under the receiver operating characteristic curve 0<middle dot>903 [95% CI 0<middle dot>795-0<middle dot>981] in YNHHS; 0<middle dot>890 [0<middle dot>839-0<middle dot>938] in MSHS for apical-4-chamber acquisitions) and transthyretin amyloid cardiomyopathy (0<middle dot>907 [0<middle dot>874-0<middle dot>932] in YNHHS; 0<middle dot>972 [0<middle dot>959-0<middle dot>983] in MSHS for parasternal acquisitions). In YNHHS, 40 (58%) of 69 hypertrophic cardiomyopathy cases and 22 (46%) of 48 transthyretin amyloid cardiomyopathy cases would have had a positive screen by AI-POCUS at a median of 2<middle dot>1 (IQR 0<middle dot>9-4<middle dot>5) years and 1<middle dot>9 (0<middle dot>6-3<middle dot>5) years before diagnosis. Moreover, among 25 261 participants without known cardiomyopathy followed up over a median of 2<middle dot>8 (1<middle dot>2-6<middle dot>4) years, AI-POCUS probabilities in the highest (vs lowest) quintile for hypertrophic cardiomyopathy and transthyretin amyloid cardiomyopathy conferred a 17% (adjusted hazard ratio 1<middle dot>17, 95% CI 1<middle dot>06-1<middle dot>29; p=0<middle dot>0022) and 32% (1<middle dot>39, 1<middle dot>19-1<middle dot>46; p<0<middle dot>0001) higher adjusted mortality risk, respectively. Interpretation We developed and validated an AI framework that enables scalable, opportunistic screening of under- recognised cardiomyopathies through simple POCUS acquisitions. Copyright (c) 2025 The Author(s). Published by Elsevier Ltd. This is an Open Access article under the CC-BY-NC-ND 4.0 license.