Akademik Veri Yönetim Sistemi
Yalçın H. Ç.(Yürütücü), Yavuz M. M., Yılmaz O.
Diğer Ülkelerdeki Kamu Kurumları Tarafından Desteklenmiş Proje, 2021 - 2025
Aortic valve defects are among the most prevalent clinical conditions. A severely damaged or non-functioning aortic valve is replaced with a mechanical heart valve (MHV) or a bioprosthetic heart valve (BHV). The former is made of highly durable pyrolytic carbon and metal alloys, while the latter is fabricated from biologically derived exogenously crosslinked soft collagenous tissue to form the leaflet biomaterials. MHVs have enhanced durability but implanted via invasive surgery and require anticoagulant therapy upon implantation. BHVs can be implanted via open heart surgery or via less invasive transcatheter aortic valve replacement (TAVR) procedure. However, these degenerate at a faster rate usually in less than 10 years. Therefore, the choice of a MHV or a BHV replacement is an important consideration, influenced by the trade-offs between the eventual need for reintervention due to BHV deterioration and the risk associated with long-term anticoagulation for the MHV. In current practice, MHVs are recommended for younger patients and BHVs are recommended for the elderly. With advancements of valve production technologies, BHVs are becoming more popular and applicable to younger and lower risk patients. According to the data from the Nationwide Inpatient Sample, approximately 50,000 aortic heart valves are replaced annually in the United States alone. While about 64% of all the aortic valve replacement surgeries in 2001 used BHV, this number increased to 82% in 2011. Similarly, total annual number of aortic valve replacements in HMC heart hospital is around 100, the majority of which are BHVs. Compared to surgical valve replacements, TAVR is traditionally applied to high risk patients. However, U.S. Food and Drug Administration recently approved the procedure for intermediate risk patients as well. With its minimally invasive character, it is expected in the near future that majority of valve replacements will be via TAVR.
For successful TAVR, it is critically important to predict the performance of BHV upon implantation. As there is no direct view or access to the affected anatomy, accurate preoperative planning is crucial for a successful outcome. The most important decisions during planning are selecting the proper implant type/size as well as positioning of the valve in the aortic root. Due to the wide variety in device sizes and types and non-circular annulus shapes, there is often no obvious choice for a given patient. Most clinicians base their final decision on their previous experience. However, it is desirable to mechanically assess selected BHV before implantation, to ensure proper valve function with minimal complications upon implantation. Mechanical assessment is being able to realistically simulate opening and closing of the valve to examine valve function. Such an assessment will prevent improper valve selection, which would lead to possible complications such as paravalvular leaks or conduction problems, major issues for TAVR. Assessment of mechanical stresses using computational modeling offers a solution that has been increasingly utilized to evaluate native and bioprosthetic valve mechanics. The technique is based on solving governing blood flow and valve displacement equations, which will enable to map blood flow through the valve and calculate hemodynamic and structural stresses acting on the valve leaflets and roots. With computational modeling, several valve function parameters such as transvalvular pressure gradient TPG, effective orifice area EOA, and leaflet coaptation surface area CSA can be calculated. In addition, hemodynamic and structural stresses acting on the valve leaflets and roots can be studied with that approach to identify high stress locations that might later lead to calcification or fatigue induced structural deterioration of BHVs. Therefore, computational modeling offers a patient-specific optimization approach for BHV selection prior to TAVR.
Even though computational simulations can effectively be used for analyzing BHVs, such patient-specific models usually require complex procedures to set up and long computing times to obtain final simulation results, preventing prompt feedback to clinicians in time-sensitive clinical applications such as TAVR. Therefore, this approach is not practical and readily applicable in clinical TAVR therapy. A potential practical and effective solution for implementing engineering analysis to mechanical assessment of BHVs is to incorporate machine learning (ML) systems to expedite and simplify the computational biomechanical analyses relevant to TAVR. Recently, ML has started impacting biomedical research and various related applications as a consequence of its aptitude for integrating vast databases, learning randomly complex relationships and incorporating expert knowledge. Particularly, deep neural networks (DNNs) have the advantage of being very flexible since they allow the predicted outputs to be related to the input features: predicted outputs are functions of hidden variables (also known as intermediate footprints, nodes or neurons). Relevant to TAVR, theoretically, DNNs can be used to relate patient specific features with optimal BHV selection and clinical outcome by surrogating for geometry and stress analysis of BHVs. To our knowledge, such a system does not exist and is the point of investigation in the current project. Here, we will develop a DNNs system for automated assessment of different BHV designs for TAVR in a patient specific manner. The ML system will first be trained on multiple computational simulation for TAVR patients with a variety of BHV virtual implantation scenarios. Training phase will ensure the DNNs system relate patient`s clinical features and selected BHV`s features for implantation with clinical outcomes in terms of hemodynamic improvement and presence/degree of complications. The final ML system will automatically assess different BHV designs to identify optimal BHV for specific clinical TAVR cases and will support clinicians in minimizing risks during TAVR therapy planning.