You are here

Computational Hemodynamics, from Scientific Computing to Clinics: Emerging Challenges in Finite Elements (and beyond)

Alessandro Veneziani, Emory University

Alessandro Reali, Universita di Pavia

In the last 30 years, the role of scientific computing in general, and computational fluid dynamics in particular in fighting cardiovascular diseases has been growing and consolidating, moving progressively from an instrument for basic research to a supporting-decision tool. This process opened new challenges in recent years, often finalized to specific and hard clinical problems, with a strong methodological component investing finite elements and other core-solver methods. Nowadays, for instance, computational fluid dynamics is a routine part of healthcare institutions, for instance in the assessment of the Fractional Flow Reserve for coronary diseases. When scientific computing needs to meet clinical timelines, the new challenges embrace different fields, with mathematical, engineering, and ethical instances. This minisymposium aims at gathering different scientists working in this fascinating field, that is introducing a new revolution in medicine (after the one brought by medical imaging).  From the methodological point of view, novel solvers with high accuracy and efficiency need to be devised at the top of basic finite element approaches, to obtain accuracy and efficiency consistent with clinical needs. The isogeometric analysis is now a consolidated tool in this field, yet other highly customized methods are emerging. The analysis and the evolution of these methods – related to medical problems -  will be considered in this minisymposium.  On the other hand, practical issues concern the endemic lack of data that make well posed the fluid-dynamics problems. Specific discussions on the different possible methods for dealing with defective boundary problems and, more in general, data assimilation procedures will be considered. Data assimilation also calls for model reduction techniques, which can make the solution of inverse problems in computational hemodynamics compatible with clinical timelines. Another topic of great interest is the uncertainty quantification when working with parametrized problems fed by measures suffering from many inaccuracies. Finally, the construction of protocols, benchmarks, and infrastructures for the design of computer-aided clinical trials will be treated. The breath of the minisymposium ranges therefore from core methodological problems to translational research where finite elements or other methods are critical for the solution of clinical problems. The final aim is the promotion of discussions and new collaborations finalized to a progressively more effective understanding and therapy of cardiovascular diseases. The support of the US NSF DMS 1620406 is greatly acknowledged.