Surgical therapy in the cardio-vascular system aims at restoring blood flow to compromized organs and tissues at physiological conditions. The goal is to avoid reverse flow and low shear areas as these are indicators for recurrent cardiovascular disease. Alternate treatments cannot be tested in the patient and since the surgeon does not have the optimal design tools he traditionally bases the therapy exclusively on medical imaging and his own experience.

In presenting an integrated software tool which assists the physician in constructing and evaluating a combined anatomic/physiologic model to assess the outcome of alternative treatment plans for an individual patient we establish a new paradigm of predictive medicine. The tool integrates medical imaging, geometric modeling, discretization, FEM-based computational fluid dynamics as well as scientific visualization.

This presentation will first summarize our work in the fields of medical imaging and image processing, geometric model generation, mesh generation, numerical simulation and visualization by using concrete physiological examples. We will then focus on the numerical method used and their optimization. The complex 3D geometry of the cardiovascular system and the pulsatile nature of blood flow require an efficient discretization method. We present a SUPG-method using linear tetrahedra for the incompressible Navier-Stokes equations. We increase the method’s efficiency by an anisotropic mesh adaptation procedure by around one order of magnitude in terms of computer time and memory requirements. We demonstrate our method using a selected number of physiologic models as for example a stenotic porcine aorta as well as a human aortic model.