Purpose

Transcatheter Aortic Valve Replacement (TAVR) has transformed the treatment of severe aortic stenosis, especially in patients at high surgical risk. There are many valves of different sizes and selecting the best one will determine the success of the procedure and long term durability.

Currently, valve selection relies heavily on clinician judgment, guided by anatomical imaging and patient demographics. However, this process is subjective and varies across institutions, introducing risk for complications such as paravalvular leak.

TAVR AI aims to bring greater precision and consistency to TAVR planning through machine learning. By training models on real-world patient data, we developed tools to assist in selecting between balloon-expandable and self-expanding valves, and to predict optimal valve size. These AI-driven models integrate complex anatomical and clinical features to support more standardized, data-driven decision-making. Ultimately improving safety, outcomes, and the efficiency of care.