CAGFound, a video-based foundation model, accurately detected severe aortic stenosis and severe mitral regurgitation from coronary angiograms, achieving an AUROC of 0.879 and 0.896 respectively in external validation.
Observational (n=12,435)
Yes
Does the CAGFound video-based foundation model accurately detect severe aortic stenosis and severe mitral regurgitation from coronary angiograms compared to other models?
A novel video-based foundation model, CAGFound, accurately screens for severe aortic stenosis and severe mitral regurgitation using standard coronary angiograms, potentially improving opportunistic detection of valvular heart disease.
Effect estimate: AUROC 0.879 (95% CI 0.809-0.948)
Abstract Coronary heart disease (CAD) is the leading cause of death worldwide, and coronary angiography (CAG) serves as the gold standard for its assessment. Valvular heart diseases, such as severe aortic stenosis (AS) and severe mitral regurgitation (MR), frequently coexist with CAD yet are often underdiagnosed. Opportunistic screening for these conditions at the time of CAG could influence therapeutic strategies and improve prognosis. This study developed and validated a foundation model for the automated screening of severe AS and severe MR from CAG videos. The study presents CAGFound, a video-based foundation model that was self-supervised pre-trained on CAG sequences from seven medical centers and subsequently adapted to two downstream tasks: screening for severe AS and severe MR. Two internal and external validation datasets were retrospectively enrolled from the First Medical Center and the Sixth Medical Center of Chinese PLA General Hospital, respectively. A total of 117,383 unlabeled CAG sequences were used to build CAGFound. For the detection of severe AS, CAGFound achieved an area under the receiver operating characteristic curve (AUROC) of 0.932 (sensitivity 0.767, specificity 0.921) on the internal test dataset and maintained robust performance on the external validation dataset, with an AUROC of 0.879 (sensitivity 0.800, specificity 0.955). For the detection of severe MR, the model demonstrated an AUROC of 0.933 (sensitivity 0.738, specificity 0.938) on the internal dataset and an AUROC of 0.896 (sensitivity 0.754, specificity 0.855) on the external cohort. The performance of CAGFound was also compared with other video-based foundation models, VideoMAEv2 and Video Swin. CAGFound achieved the highest AUROC and demonstrated the best calibration performance (Brier score 0.122, R 2 0.478) compared with VideoMAEv2 (Brier score 0.159, R 2 0.306) and Video Swin (Brier score 0.162, R 2 0.306). CAGFound enables accurate, automated screening for severe AS and severe MR during CAG. It has the potential to increase detection rates, facilitate timely clinical referral, and improve prognosis without requiring additional contrast administration or procedures.
Recent publication with AI in cardiology buzz; potential practice change for screening.
Zhang et al. (Wed,) conducted a observational in Severe aortic stenosis and severe mitral regurgitation (n=12,435). CAGFound (video-based foundation model) vs. Transthoracic echocardiography (reference standard) was evaluated on Detection of severe aortic stenosis (AUROC in external validation) (AUROC 0.879, 95% CI 0.809-0.948). CAGFound, a video-based foundation model, accurately detected severe aortic stenosis and severe mitral regurgitation from coronary angiograms, achieving an AUROC of 0.879 and 0.896 respectively in external validation.