Periodic Reporting for period 1 - OPTAVI (Adjoint-based OPTimization of deflection-based cerebral embolic protection devices for reducing stroke risk in Transcatheter Aortic Valve Implantation)
Reporting period: 2021-10-01 to 2023-09-30
- The problem/issue addressed:
The project OPTAVI aimed at developing optimization strategies based on gradient-based techniques widely used in fluid mechanics, for enabling improvement of transcatheter prosthetic heart valve implantation (TAVI) procedures. The project relied on developing sophisticated image-based blood flow simulation tools, which were to be designed to leverage the power of advanced hybrid-node (GPU-accelerated) supercomputers. The latter serves as a workhorse tool for evidence generation on fluid dynamics aspects of TAVI operation, and also as a base tool for optimization procedure.
- Why is it important for society
TAVI operations have grown to be the procedure of choice for a wide variety of aortic valve disease patients, mainly those at older age groups. This procedures are favourable because they eliminate the need for open heart surgery, which may impose high risks on a significant number of patients, especially those with other co-existing cardiac diseases. TAVI allows the impaired heart valve to be replaced via a catheter through an artery, which minimizes the hospitalization as well. It however, comes with certain risks, such as a high incidence of peri- and postoperative stroke. The "shoving" motion of the packaged heart valve through the arteries (which are likely to be plaqued in AV patients), and parachute-like opening of the valve inside the diseased valve site (which is often calcified) are prone to release a significant amount of plaque debris into blood flow. These released objects may then travel to the brain, interrup the cerbral blood flow circulation, and eventually cause stroke. In addition, TAVI is associated with stroke risk factors post-operation, that is, after the deployment of the valve. The latter is mainly due to the design of the prosthetic valve which can lead to narrow jets of blood flow that are likely to generate blood clots due to strong shear forces. Improving the design of TAVI and its implantation procedure can reduce the risk of stroke in TAVI, which reduces healthcare costs (higher durabiliy and lower need for re-operation) and improves the quality of life for the recipients of these valves (it reduces the morbidity factors associated with an impaired cerebral blood flow, e.g. dementia or small strokes or mortality factors such as fatal strokes). These improvements yield significant socioeconomical values.
- What are the overal objectives:
The objective of the project was to investigate the pacticality of an adjoint-based optimization procedure for a TAVI design. It relied on developing computational fluid dynamics tools for blood flow analysis and adjoint-based methods on top of the simulation environment, with the goal of minimizing the thrombogenicity of a aortic valve design.
The project OPTAVI aimed at developing optimization strategies based on gradient-based techniques widely used in fluid mechanics, for enabling improvement of transcatheter prosthetic heart valve implantation (TAVI) procedures. The project relied on developing sophisticated image-based blood flow simulation tools, which were to be designed to leverage the power of advanced hybrid-node (GPU-accelerated) supercomputers. The latter serves as a workhorse tool for evidence generation on fluid dynamics aspects of TAVI operation, and also as a base tool for optimization procedure.
- Why is it important for society
TAVI operations have grown to be the procedure of choice for a wide variety of aortic valve disease patients, mainly those at older age groups. This procedures are favourable because they eliminate the need for open heart surgery, which may impose high risks on a significant number of patients, especially those with other co-existing cardiac diseases. TAVI allows the impaired heart valve to be replaced via a catheter through an artery, which minimizes the hospitalization as well. It however, comes with certain risks, such as a high incidence of peri- and postoperative stroke. The "shoving" motion of the packaged heart valve through the arteries (which are likely to be plaqued in AV patients), and parachute-like opening of the valve inside the diseased valve site (which is often calcified) are prone to release a significant amount of plaque debris into blood flow. These released objects may then travel to the brain, interrup the cerbral blood flow circulation, and eventually cause stroke. In addition, TAVI is associated with stroke risk factors post-operation, that is, after the deployment of the valve. The latter is mainly due to the design of the prosthetic valve which can lead to narrow jets of blood flow that are likely to generate blood clots due to strong shear forces. Improving the design of TAVI and its implantation procedure can reduce the risk of stroke in TAVI, which reduces healthcare costs (higher durabiliy and lower need for re-operation) and improves the quality of life for the recipients of these valves (it reduces the morbidity factors associated with an impaired cerebral blood flow, e.g. dementia or small strokes or mortality factors such as fatal strokes). These improvements yield significant socioeconomical values.
- What are the overal objectives:
The objective of the project was to investigate the pacticality of an adjoint-based optimization procedure for a TAVI design. It relied on developing computational fluid dynamics tools for blood flow analysis and adjoint-based methods on top of the simulation environment, with the goal of minimizing the thrombogenicity of a aortic valve design.
The work performed during the project is described in three categories.
1. Development of a 3D imaged-based computational fluid dynamics tool for high-resolution simulation of blood flow in patient-specific aortic geometries
Clinical guidelines on TAVI suggest that the hemodynamic performance of the valve is tightly linked with the morphology of the aorta, where the valve is to be implanted. The curvature of the aorta is specific for each patient, and oftentimes is already affected by the aortic valve disease (for instance, bicuspid valve patients are known for a high incidence of aortic dilation). This led us to develop and couple an image-based arterial geometry analysis tool with our legacy GPU-accelerated solver, aiming to incorporate detailed models of patient-specific aortic geometries. The resulting tool takes CT scans of the patients chests, creates suitable input models for integration into blood flow simulation tool and finally performs detailed blood flow simulations that are highly resolved in time and space. The base solver was designed to capture blood flow details such as turbulent shear stresses with high accuracy, enabling more precise predictions and laying the groundwork for more rigorous optimizations. The results produced by our flow simulation tool packed in extremely large datasets of sizes beyond 10TBs for each beat cycle, which provided unprecedented spatiotemporal detayils into the flow phenomena in the aorta.
2. For the adjoint-based methodology, we started by a simple valve design, given the complexity of the full problem in hand. Adjoint-based methods are costly in nature, given that they require several iterations of a so-called forward-backward looping procedure. Given that 3D simulations are significantly costly even using the most advanced GPUs (a single heart beat simulation required three days allocation on eight powerful P100 GPUs), performing such simulations is not practical in a repetitive fashion that is required for optimization. Therefore, we developed an optimization code based on a simplified 2D design of a prosthetic valve. Leveraging both linear and nonlinear optimization techniques which could utilize the optimization code for both route-cause identification as well as control of adverse blood flow events. Particularly, our optimization framwork took a certain geometrical mask as a user input, and could identify the most relevant root causes for an unphysiological flow (e.g. highly turbulent) to develop within the mask. In the next step, and aiming to extend our methodology to 3D environments, we explored the use of artifical neural networks together with our direct numerical simulations. In particular, we addressed an specifc type of these techniques, namely, the Physics-informed neural networks (PINNs). The PINNs approach can be challenging for unsteady flow scenarios and high geometrical details, and proved to be costly to train even in a 2D setting. Future advancements in this area could potentially be highly relevant for the optimization procedures in 3D settings.
1. Development of a 3D imaged-based computational fluid dynamics tool for high-resolution simulation of blood flow in patient-specific aortic geometries
Clinical guidelines on TAVI suggest that the hemodynamic performance of the valve is tightly linked with the morphology of the aorta, where the valve is to be implanted. The curvature of the aorta is specific for each patient, and oftentimes is already affected by the aortic valve disease (for instance, bicuspid valve patients are known for a high incidence of aortic dilation). This led us to develop and couple an image-based arterial geometry analysis tool with our legacy GPU-accelerated solver, aiming to incorporate detailed models of patient-specific aortic geometries. The resulting tool takes CT scans of the patients chests, creates suitable input models for integration into blood flow simulation tool and finally performs detailed blood flow simulations that are highly resolved in time and space. The base solver was designed to capture blood flow details such as turbulent shear stresses with high accuracy, enabling more precise predictions and laying the groundwork for more rigorous optimizations. The results produced by our flow simulation tool packed in extremely large datasets of sizes beyond 10TBs for each beat cycle, which provided unprecedented spatiotemporal detayils into the flow phenomena in the aorta.
2. For the adjoint-based methodology, we started by a simple valve design, given the complexity of the full problem in hand. Adjoint-based methods are costly in nature, given that they require several iterations of a so-called forward-backward looping procedure. Given that 3D simulations are significantly costly even using the most advanced GPUs (a single heart beat simulation required three days allocation on eight powerful P100 GPUs), performing such simulations is not practical in a repetitive fashion that is required for optimization. Therefore, we developed an optimization code based on a simplified 2D design of a prosthetic valve. Leveraging both linear and nonlinear optimization techniques which could utilize the optimization code for both route-cause identification as well as control of adverse blood flow events. Particularly, our optimization framwork took a certain geometrical mask as a user input, and could identify the most relevant root causes for an unphysiological flow (e.g. highly turbulent) to develop within the mask. In the next step, and aiming to extend our methodology to 3D environments, we explored the use of artifical neural networks together with our direct numerical simulations. In particular, we addressed an specifc type of these techniques, namely, the Physics-informed neural networks (PINNs). The PINNs approach can be challenging for unsteady flow scenarios and high geometrical details, and proved to be costly to train even in a 2D setting. Future advancements in this area could potentially be highly relevant for the optimization procedures in 3D settings.
Even though the optmization problem tackled remains unsolved due to an extemely challenging nature, our simulation tool which is a by-product of the project provides a seemless image-based CFD tool for aortic blood flow analysis in clinical environments. Offering a flexible and relatively easy-to-use image analysis interface, and also owing to advanced high-performance computing (HPC) capabilities, our solver can perform an aortic blood flow simulation over a full beat cycle in a weak, using only 8 GPUs (this amount of HPC resources can be found reliably on majority of university of hospital clusters). The 7 days period includes preprocessing (creating the geometric model and BCs), running simulation, and postprocessing (creating snapshots and animated pictures of the flow).