Biomedical Flows and Structures

The purpose of this project is to conduct clinically relevant research in a few selected medical problems where computational fluid and solid dynamics seem to have great potential for increased understanding and change of clinical treatment. In each of the selected problems, we have close collaborations with top national or international clinical researchers. The collaborations have proven successful and effective, and we have already published multiple papers in prestigious medical journals. In addition, the researchers in the project have published and will continue to publish results from computational advances relevant to the medical applications. These results are of particular importance for using more advanced models in the future.

The bioflow projects outlined in the original proposal have matured significantly, and new projects have been started where we have seen excellent partners and research ideas. The plan for future research presented here is therefore a significantly revised version of the original plans from 2006. The following subprojects are now in focus:

1. Aneurysm Formation and Rupture in the Circle of Willis
2. Cerebrospinal Fluid Flow and Cyst Formation in the Spinal Chord
3. Pressure-Stress Conditions in the Brain under NPH
4. Functioning of the Mitral Valve

Many computational fluid dynamics groups addressing questions similar to ours have a tendency to use very advanced and computationally intensive models. We aim to excel in picking a model that represents the right balance between being able to catch the main physics of the medical problem and being sufficiently fast to run so that extensive experiments can be conducted. Quantitative, statistical tools from the previously mentioned Model Calibration project will be essential for choosing the right model complexity and identifying input parameters from medical images and flow measurements.

Contrary to many other groups targeting blood flow computations, we have so far not implemented a strong focus on interaction between blood flow and flexible vessel walls, which result in computationally very expensive FSI models. Instead, laboratory experiments, in vivo measurements, and simulations point to the importance of localized turbulence and patient-specific geometries for increased medical understanding. Future comparison with full FSI models is obviously needed to verify this assumption. However, at present it is difficult to achieve sufficient medical progress in biomedical flows with time-consuming FSI models. Successful research on FSI in the Robust Solvers project may produce sufficiently fast FSI solvers for future large-scale medical investigations.

Using computational experiments, our research strategy is to provide evidence for accepting or rejecting precise hypotheses stated by medical researchers and clinicians. We xpect such results to be of high relevance for medical journals. The real potential of simulations in medical practice will be investigated through comprehensive patient studies. This is normally very labor consuming, but our strong competence in scripting techniques in combination with FEniCS solvers allow for tailored problem solving environments with a maximum degree of automation in running patient studies.

Since we aim at understanding and improving treatment of conditions that are associated with severe health problems and substantial costs, there is naturally a potential for commercialization and media coverage if our approach and research are successful.