Aniket Venkatesh

BioE Ph.D. Proposal Presentation

Date and Time: Thursday, January 22nd, 2026, at 1:00 PM (EST)

Location: TEP 216E 

Advisor: 

Lakshmi Prasad Dasi, Ph.D. (Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University)

Committee:

John Oshinski, Ph.D. (Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University)

Rudolph Gleason, Ph.D. (Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University)

Hanjoong Jo, Ph.D. (Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University)

Vinod Thourani, M.D. (Piedmont Heart Institute, Piedmont Healthcare)

 Predicting Transcatheter Aortic Valve Leaflet Thrombosis Risk Using Pre-Procedural Computational Modeling

    Aortic valve (AV) stenosis is the most common heart valve disease, present in about 30% of patients aged 65 or more and exacerbated by the congenitally abnormal bicuspid AV. Currently, the only treatment is an aortic valve replacement, which is commonly performed via transcatheter aortic valve (TAV) replacement (TAVR), a minimally invasive balloon-expandable (BE) or self-expandable (SE) procedure for patients at high surgical risk. However, the formation of blood clots on the TAV leaflets, or leaflet thrombosis (LT), is one of the most common adverse outcomes post-TAVR. About 15% of all patients who undergo TAVR develop LT, which increases the risk of stroke and early valve deterioration. Currently, LT is only diagnosed post-procedurally through detection of hypoattenuated leaflet thickening (HALT) in cardiac computed tomography (CT) scans taken days to months after procedure. Prior studies have examined direct correlations between various post-TAVR geometric and hemodynamic features on HALT risk, but no standardized methods have been developed to predict HALT pre-procedurally in the clinic. Hence, there is an urgent unmet need for rapid prediction of patient-specific post-TAVR HALT to aid in procedural planning. 

    So far, the project has resulted in a preliminary computational pipeline that predicts some post-TAVR geometric and hemodynamic features from pre-TAVR CT of tricuspid AV patients, but additional work is needed to further decrease the computational time and cost of these predictions, while incorporating additional TAVR models and bicuspid AVs. Therefore, the goal is to develop, validate, and test a robust computational pipeline that can predict patient-specific HALT risk after BE and SE TAVR in patients with tricuspid and bicuspid AVs, using only pre-procedural imaging. To do so, the following aims will be addressed: 1) Develop and validate a quick-response simulation-guided computational pipeline to pre-operatively predict post-TAVR geometric and hemodynamic features, 2) Predict HALT risk in patients following BE TAVR, and 3) Develop and test metrics to predict HALT risk following SE TAVR. By addressing one of the most common TAVR complications and methods to predict it while minimizing computational time and cost, this work will contribute to the development and validation of a rapidly deployable TAVR procedural planning tool that can be used in the clinic.