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

Abstract

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.

Automated, standardized, quantitative analysis of cardiovascular borders on chest X-rays using deep learning for assessing cardiovascular disease

“…[1][2][3] Recent studies have also demonstrated AI's potential in cardiovascular disease for diagnosing heart failure, predicting cardiovascular disease risks, and identifying various types of valvular diseases using CXRs. [4][5][6][7][8] AI systems trained to predict cardiovascular abnormalities in CXRs can provide saliency maps for their explainability, which highlight the areas focused on making diagnoses. 5 7 However, it is important to note that these heatmaps might have limitations, particularly in pinpointing specific abnormalities or diagnosing rare diseases.…”

“…The use of "end-to-end" supervised learning, where AI directly learns from CXRs with abnormalities compared to a control group, is a widely adopted approach in current AI research. This method has been extensively applied in the field of cardiovascular disease to predict conditions such as acute chest pain syndrome 24 , aortic dissection 25 , LV systolic dysfunction 6 , structural LV disease 7 , valvular heart disease 5 , aortic stenosis 26 , and atrial fibrillation 27 using CXRs. Other studies have also tried to predict the 10-year risk for major adverse cardiovascular events using CXRs.…”