Cardiovascular Imaging, Modeling, and Simulation

2 Apr 2010

Project SIMCARD involves large-scale computer simulations and a close connection between mathematics, scientific computing, and clinical experiments, to work on medical imaging and numerical simulations of complex bioengineering problems.

by Adélia Sequeira

SIMCARD is an interdisciplinary project that develops, analyzes, and simulates mathematical models of the cardiovascular system. It involves large-scale computer simulations and demands a collective expertise. It requires a close connection between mathematics, scientific computing, and clinical experiments, to work on medical imaging and numerical simulations of complex bioengineering problems.

Project SIMCARD formally started in early 2009. It brings together expertise from Portuguese universities (IST, FEUP, FML) and The University of Texas at Austin (ICES research teams led by Chandrajit Bajaj and Tom J.R. Hughes). 

The first stint focuses on medical image analysis, virtual model reconstruction, and patient-specific CFD simulations of cerebral aneurysms and fluid-structure interaction. The project joined mathematicians and clinicians, thereby broadening the understanding of each field's goals, difficulties, and potentials.

One of the greatest challenges of performing simulations from clinical data is the inherent unquantifiable error that is carried out from the earliest stage. Noise and flow artifacts in the medical images and clinical measurements lead to geometric models that contain some uncertainty. Patient-specific case studies to quantify relative bounds to this uncertainty uncovered a dire need for care in the virtual model reconstruction from medical images. In certain cases, relative errors in the computed flow field due to geometric uncertainty can be of the order of 50 percent. These can be further augmented by combining uncertainty in the choice of appropriate constitutive models for the blood and the boundary conditions. In the case of aneurysms, the uncertainty due to the virtual model definition is comparable to that when choosing different models for the non-Newtonian behavior of blood. Short-term advances in this topic include the formation of a database of patient-specific studies and fluid-structure studies on aneurysms. A postdoctoral position is currently open for the duration of one year in conjunction with this project.

SIMCARD is funded by the Portuguese Foundation for Science and Technology (FCT) under the 2008 R&D call of the UT Austin|Portugal Program (project number UTAustin/CA/0047/2008).

Read more about CoLab research projects