YALES2BIO : outil de simulation hmodynamique S. Mendez CNRS et I3M, - PowerPoint PPT Presentation

YALES2BIO : outil de simulation hémodynamique S. Mendez CNRS et I3M, Université Montpellier II Avec C. Chnafa, E. Gibaud, J. Sigüenza, V. Zmijanovic and F. Nicoud Institut de Mathématiques et de Modélisation de Montpellier Journée de recherche translationnelle sur les systèmes Biomédicaux Cardio-Vasculaires – BioDev. 20/11/14

Contributors and collaborators I3M (Montpellier) CHU Toulouse • • - Christophe Chnafa - Ramiro Moreno - Etienne Gibaud - Hervé Rousseau - Marco Martins Afonso CHU Montpellier et IRRAS Technology • - Simon Mendez - Vincent Costalat - Franck Nicoud - Mathieu Sanchez - Julien Sigüenza - Vladeta Zmijanovic CHU Nîmes • - Iris Schüster CORIA (Rouen) • - Ghislain Lartigue • Lab. Pharm-Ecologie Cardiovasculaire - Vincent Moureau (Avignon) - Claire Maufrais LMGC (Montpellier) • - Stéphane Nottin - Dominique Ambard - Frédéric Dubois Horiba Medical (Montpellier) • - Franck Jourdan - Damien Isèbe - Rémy Mozul • L2C (Montpellier) - Manouk Abkarian 2

Numerical simulations of blood flows Numerical simulations = Solving the equations by computers Ex: fluid equations ∂ U i = 0 ∂ x i   ρ dU i i − ∂ P µ ∂ U i + ∂ = ρ F     dt ∂ x i ∂ x j ∂ x j   solved by finite elements, finite volumes,… 3

Why numerical simulations? To analyze Data correlation Flow description Chnafa et al. 2014 Torii et al. 2009 Sanchez et al. 2014 To predict Surgical outcome Device design Van Caneeyt et al. 2009 Marsden et al. 2009 4

Analyzing with numerical simulations Sanchez et al. 2014: Torii et al. 2009: infer the aneurysm wall properties how aneurysm shape influences flow in aneurysms => Voir V. Costalat 15h55 5

Predicting with numerical simulations Van Canneyt et al. 2013: Design of arteriovenous graft => Voir P. Verdonck 14h15 Marsden et al. 2009: Design of Fontan graft 6

YALES2BIO • I3M is developing YALES2BIO • From existing solver YALES2 (CORIA) http://www.math.univ-montp2.fr/~yales2bio Method: see - Moureau et al. 2011, - Chnafa et al. 2014a & 2014b, - Mendez et al. 2014 • Two applications : 1. Left heart flow from medical images 2. Artificial heart flow 7

The heart. In brief AG AD VG VD 8

Heart flow is extremely interesting • For fundamental fluid mechanics • For medical reasons: - Wall shear stress - High pressure - Diagnosis (aortic valve stenosis) Markl et al. 2013 Kilner et al. 2000 • For future artificial devices (Carmat, Abiocor, Reinheart,…) 9

What should we predict? • Flow related to heart movement • Wall movement prediction => errors • For flow predictions only, use the medical images 10

Image registration algorithm From images (color), 11

Image registration algorithm From images (color), 12

Image registration algorithm From images (color), ⇒ 4D mesh of the flow domain ⇒ Flow computation in a domain with prescribed motion Moreno et al. 2008, Midulla et al. 2012, Chnafa et al. 2014 13

Animation of the flow - Grid 3 millions of tetrahedra - 70 cycles computed - 1 cycle = 4h on 60 processors 14

Strong cycle to cycle variations Recently : Zajac et al. 2014. Turbulence intensity as a biomarker? 15

Modeling the transport in the heart 0.5 s Time spent in the ventricle 0 s 0.5 s Time spent in the ventricle 0 s For more details: PhD defense of C. Chnafa, tomorrow 14h 16

Total artificial hearts Syncardia heart (Slepian et al., J. Biomech. 2013) Abiocar, Carmat hearts,… Carmat Abiocor How to compute the flow in these systems? 17

Principle of the CARMAT artificial heart Right ventricle Left ventricle Secondary fluid Pump Schematic from the French newspaper ‘Le Monde’ 18

Accounting for immersed structures Flexible membranes and valves 19

Fluid-structure interaction Coupling with LMGC90 (with Ambard, Dubois, Jourdan, Mozul, Sigüenza) Forces � Displacements 20

Example with valves Thesis : J. Sigüenza 21

Back to the artificial heart A first attempt in half the system: unstructured LES + flexible membrane 22

Configuration • A domain mimicking the left heart flow (only half of the heart is considered). From lungs To aorta Blood Side view membrane Top view Secondary fluid 23

Movie presentation 24

Movie (5 cycles from the start) 25

Residence times movie (8 cycles) 26

The artificial heart case 1. Benefits from the experience on physiological heart 2. Industrial applications for Fluid-Structure coupling 3. Proof of concept case => application on real geometries soon 27

Computations for industrial applications The use of a high-fidelity solver Artificial hearts Blood analyzers 28

Counting and sizing with the Coulter effect Region of interest Velocity 6 m/s in a 50 µ m aperture 29

Passage of red blood cells in the orifice Thesis : E. Gibaud Method : Relation between the red blood cell dynamics and measured volume Objective : minimize errors by device optimization 30

Thank you for your attention • I3M (Montpellier) • CHU Montpellier et IRRAS Technology - Christophe Chnafa - Vincent Costalat - Etienne Gibaud - Mathieu Sanchez - Marco Martins Afonso - Simon Mendez • CHU Toulouse - Franck Nicoud - Ramiro Moreno - Julien Sigüenza - Hervé Rousseau - Vladeta Zmijanovic • Lab. Pharm-Ecologie CORIA (Rouen) • Cardiovasculaire (Avignon) - Ghislain Lartigue - Claire Maufrais - Vincent Moureau - Stéphane Nottin • LMGC (Montpellier) • L2C (Montpellier) - Dominique Ambard - Manouk Abkarian - Frédéric Dubois Horiba Medical (Montpellier) • - Franck Jourdan - Damien Isèbe - Rémy Mozul http://www.math.univ-montp2.fr/~yales2bio • CHU Nîmes - Iris Schüster We thank for financial support: