Study of an Artificial Intelligence Model to Identify Severe Low-Gradient Aortic Stenosis Using Echocardiogram Images

What this Trial Is About

Researchers are evaluating an artificial intelligence AI model designed to identify severe low-gradient aortic valve stenosis using echocardiographic images. This condition requires accurate diagnosis to determine the need for surgical treatment. The study is observational and aims to validate the AI models diagnostic accuracy by comparing its assessments with those of experienced physicians in real-world practice. Participants with suspected significant low-gradient aortic stenosis will have standard transthoracic echocardiography performed as part of their clinical care. The echocardiographic images will then be analyzed by the AI model to classify the stenosis as severe or moderate. Additional diagnostic tests may be done if clinically needed, but the AI results will not affect clinical decisions. Approximately 300 patients will be enrolled during 2025-2026. During the study, the AI models performance will be measured by the area under the receiver operating characteristic curve AUC and diagnostic accuracy in specific patient subgroups. These assessments will occur around the time of the nearest Heart Team meeting following echocardiography, typically within one week. The study involves no additional risk or changes to patient care, focusing on collecting data to improve future diagnostic tools.

AI Assessment of Low-Gradient Aortic Stenosis Severity Based on Echocardiography