PASCA²L: PAtient Specific Computational Analysis of Aortic hemodynamics over Large data sets
The project aims at developing a computational framework exploiting Biomechanics computational methods, such as CFD, for the study of aortic flow in physiological, pathological and postoperative using clinical data sets having considerable size (and measurements), from which are reconstructed patient-specific vascular models. The large volume of data that we intend to treat raises two important technological challenges: 1) the simulated patient-specific aortic arch requires significant computing resources to meet the requirements of accuracy; 2) the numerical problems in this field generally require the solution of linear systems of several millions of equations to be solved on time intervals that cover some cardiac cycles (i.e., thousands of iterations).
In order to achieve the project, we use the computational fluid dynamic code LifeV, which is a finite element library, designed for massive parallel simulations with state of the art techniques for biomedical flow simulations. The code is a joint collaboration between École Polytechnique Fédérale de Lausanne in Switzerland, Politecnico di Milano in Italy, INRIA in France and Emory University in the USA.
The simulations are performed on the supercomputer FERMI, hosted at CINECA. FERMI is a BlueGeneQ machine designed by IBM and it is ranked 9th (November 2012) on the TOP500.