BioE PhD Proposal Presentation- Sri Krishna Sivakumar

Advisor: Lakshmi Dasi, Ph.D., School of Biomedical Engineering, Georgia Institute of Technology  

Committee Members:   

Ajit P. Yoganathan, Ph.D., School of Biomedical Engineering, Georgia Institute of Technology  

John Oshinski, Ph.D., School of Biomedical Engineering, Georgia Institute of Technology and Emory University  

Brandon Dixon, Ph.D., School of Mechanical Engineering, Georgia Institute of Technology  

Vinod H. Thourani, M.D., Department of Cardiovascular Surgery, Piedmont Heart Institute  

Predictive Computational Modeling of Transcatheter Aortic Valve Replacement

Aortic valve stenosis (AS) is a disease caused by valve degeneration, most commonly due to calcific aortic valve disease that affects 3% of all adults over 65 years of age. Based on severity, the aortic valve is replaced by a bioprosthetic aortic valve, either surgically or using a transcatheter approach. Currently, transcatheter based approaches to aortic valve replacement (TAVR) are being widely adopted, especially in patients who are at increased risk of mortality with conventional open-heart surgery. However, adverse procedural complications such as coronary obstruction and aortic root rupture can severely impact the success of the procedure. Despite the low incidence of such adverse outcomes after TAVR, they can present high mortality rates of up to 40% at 30-day follow-up. The biomechanics of TAVR complications are not fully understood. Pre-procedural cardiac computed tomography (CT) imaging is often insufficient in visualizing the complex interactions between the bioprosthetic stent and diseased aortic valve and therefore, reliable prediction of occurrence of these complications based on CT measurements remains a challenge. The goal of the proposed research is to develop a patient specific computational framework for TAVR, that is validated using post-procedural clinical imaging data, for use in risk assessment & planning to improve patient selection for TAVR. Finite element methods and computational fluid dynamics will be used to create the predictive models for simulating TAVR in patient specific geometries and for quantitative risk assessment of coronary obstruction. An in vitro flow simulator that allows for reliable reproduction of in vivo conditions in patient specific 3D printed geometries, as well as retrospective procedural outcomes, will be used to validate results from computational modeling and improve the robustness of the prediction. The integration of computational modeling in procedural planning could be the next major step towards reducing rates of complications and maximizing the success of TAVR.