Deep learning to detect left ventricular structural abnormalities in chest X-rays.

Autor:	Bhave S; Division of Cardiology and Department of Biomedical Informatics, Columbia University Irving Medical Center, 622 West 168th Street, PH20, NewYork, NY 10032, USA., Rodriguez V; Division of Cardiology and Department of Biomedical Informatics, Columbia University Irving Medical Center, 622 West 168th Street, PH20, NewYork, NY 10032, USA., Poterucha T; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Mutasa S; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Aberle D; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Capaccione KM; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Chen Y; Inova Fairfax Hospital Imaging Center, Inova Fairfax Medical Campus, Falls Church, VA, USA., Dsouza B; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Dumeer S; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Goldstein J; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Hodes A; Hackensack Radiology Group, Hackensack Meridian School of Medicine, Nutley, NJ, USA., Leb J; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Lungren M; Department of Radiology, University of California, SanFrancisco, CA, USA., Miller M; Hackensack Radiology Group, Hackensack Meridian School of Medicine, Nutley, NJ, USA., Monoky D; Hackensack Radiology Group, Hackensack Meridian School of Medicine, Nutley, NJ, USA., Navot B; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Wattamwar K; Division of Vascular and Interventional Radiology, Department of Radiology, Montefiore Medical Center, Bronx, NY, USA., Wattamwar A; Hackensack Radiology Group, Hackensack Meridian School of Medicine, Nutley, NJ, USA., Clerkin K; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Ouyang D; Smidt Heart Institute, Cedars-Sinai Medical Center, Los Angeles, CA, USA., Ashley E; Stanford Center for Inherited Cardiovascular Disease, Stanford University School of Medicine, Palo Alto, CA, USA., Topkara VK; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Maurer M; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Einstein AJ; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA.; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Uriel N; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Homma S; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Schwartz A; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA., Jaramillo D; Department of Radiology, Columbia University Irving Medical Center, NewYork, NY, USA., Perotte AJ; Division of Cardiology and Department of Biomedical Informatics, Columbia University Irving Medical Center, 622 West 168th Street, PH20, NewYork, NY 10032, USA., Elias P; Division of Cardiology and Department of Biomedical Informatics, Columbia University Irving Medical Center, 622 West 168th Street, PH20, NewYork, NY 10032, USA.; Seymour, Paul, and Gloria Milstein Division of Cardiology, Department of Medicine, Columbia University Irving Medical Center, NewYork-Presbyterian Hospital, 630 West 168th Street, NewYork, NY 10032, USA.
Jazyk:	angličtina
Zdroj:	European heart journal [Eur Heart J] 2024 Jun 07; Vol. 45 (22), pp. 2002-2012.
DOI:	10.1093/eurheartj/ehad782
Abstrakt:	Background and Aims: Early identification of cardiac structural abnormalities indicative of heart failure is crucial to improving patient outcomes. Chest X-rays (CXRs) are routinely conducted on a broad population of patients, presenting an opportunity to build scalable screening tools for structural abnormalities indicative of Stage B or worse heart failure with deep learning methods. In this study, a model was developed to identify severe left ventricular hypertrophy (SLVH) and dilated left ventricle (DLV) using CXRs. Methods: A total of 71 589 unique CXRs from 24 689 different patients completed within 1 year of echocardiograms were identified. Labels for SLVH, DLV, and a composite label indicating the presence of either were extracted from echocardiograms. A deep learning model was developed and evaluated using area under the receiver operating characteristic curve (AUROC). Performance was additionally validated on 8003 CXRs from an external site and compared against visual assessment by 15 board-certified radiologists. Results: The model yielded an AUROC of 0.79 (0.76-0.81) for SLVH, 0.80 (0.77-0.84) for DLV, and 0.80 (0.78-0.83) for the composite label, with similar performance on an external data set. The model outperformed all 15 individual radiologists for predicting the composite label and achieved a sensitivity of 71% vs. 66% against the consensus vote across all radiologists at a fixed specificity of 73%. Conclusions: Deep learning analysis of CXRs can accurately detect the presence of certain structural abnormalities and may be useful in early identification of patients with LV hypertrophy and dilation. As a resource to promote further innovation, 71 589 CXRs with adjoining echocardiographic labels have been made publicly available. (© The Author(s) 2024. Published by Oxford University Press on behalf of the European Society of Cardiology.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu