Project SENSUS

Based on recent works in cardiovascular fluid-structure interaction (see Fig.) and adjoint shape optimization, this project aims at the optimization of artificial blood vessels.
In particular, the shape of the blood vessels shall be improved with respect to the hemodynamic, i.e. the blood flow.
In a first step, the existing modeling and simulation framework is further developed in terms of efficiency.
Among other measures, the structural solver will be changed to a newer version that allows for adaptive mesh refinement.
In the context of the partitioned solution approach followed to solve the strongly coupled FSI problem, this demands for a change in the mapping technique that is needed to transfer fluid loads and structural displacement from one discretization to the other.

In a next step, the adjoint equations for particular cost functions shall be derived, implemented and tested.
Typically, wall-shear-stress (WSS) related quantities are taken as a quality measure for the hemodynamics in artificial blood vessels.
Besides the norm of the WSS, which should be within the pyhsiolically expected range, its gradient (WSSG) and the oscillating shear index (OSI), which quantifiess how mus the WSS change direction over one cardiac cycle, shall be considered in the cost function.
Due to the similarity of the adjoint problem to the primal FSI problem, the same solvers and and coupling procedures can be used to solve it in order to obtain a surface sensitivity.
However, the specific parameters for the overall solution methods that yield the most efficient simulations have to be found.
That means, that different choices for the temporal integration method within the field solvers as well as for the building blocks of the coupling algorithm (predictor, convergence acceleration method, sub-stepping procedure, etc.) have to be compared.

Finally, the adjoint shape optimization procedure shall be applied to a clinically relevant exemplary cases.
A first choice for such a case is the distal (downstream) anastomosis (connection) of a bypass-graft and an artery.
These connections are prone to developing a stenosis (narrowing of the vessel walls) in the anastomosis region, which may lead to a failure of the bypass graft and the need for an anew surgery.
Several studies have connected the failure rate of these anastomoses to the prevailing hemodynamics, especially to the mentioned WSS related quantities.
The adjoint shape optimization method from this project can be used to find alternative anasotmosis shapes that lead to more physiological hemodynamics.

Contact: Lars Radtke