LV Mechanics Challenge LV Mechanics Challenge

Understanding the mechanical behaviour of the heart is important in the evaluation of cardiac disease. In particular, patients with heart failure can present with a spectrum of symptoms from preserved to reduced ejection fraction. Currently it is difficult to determine the passive stiffness properties as well as the active tension development during systole. Recently, a number of methods have been proposed for utilising image information to reverse engineer the mechanical properties of the heart. These take the form of mathematical simulations of the cardiac cycle. However, there has been no objective comparison of the characteristics of these approaches.

The objective of this challenge is to compare the behaviour of different methods used to simulate the systolic and diastolic mechanics of the left ventricle. High resolution MRI data (see the animation) from four normal dogs, contributed from the National Institutes of Health, will be made available to participants [1]. The data include mesh point clouds and binary masks defining the LV geometry, and muscle fibre orientations derived from ex vivo diffusion tensor MRI. The DTI data have already been registered to the in-vivo geometry. Geometries at three states in the cardiac cycle will be given: stress free (diastasis), end of inflation (end-diastole) and end of contraction (end-systole). The geometries only reflect the location of the epicardial and endocardial surfaces- they do not encode material point displacements. In-vivo left ventricular pressures will be given for each of these three states.

Participants can use either the point cloud or binary masks to generate meshes. A hexahedral mesh will be provided. The first principal eigenvectors of the DTI data (myocardial fibre orientations) will be provided at a regularly sampled grid of points. The displacements of four points at the base of the LV will be given to provide a common reference to account for global motion. For simplicity, activation will be assumed to be simultaneous throughout the LV. Participants will be free to individualise other boundary constraints, interpolation schemes, material relations, etc. Note that non-physical "image-based" driving forces should not be used.

Model validation will be made against displacement from high resolution tagged MRI, which will not be distributed to the participants. Participants will be asked to use their individualised biomechanical models to interpolate the displacements of the same set of material points at which the fibre vectors were provided. These displacement data will be compared with data derived from the MR tagging studies (see the movie).

For further information please contact Alistair Young: a.young@auckland.ac.nz

Reference:

  1. Wang VY, Lam HI, Ennis DB, Cowan BR, Young AA, Nash MP. Modelling passive diastolic mechanics with quantitative MRI of cardiac structure and function. Med Image Anal. 2009;13(5):773-784.
Organisers Organisers
  1. Alistair Young
    (University of Auckland, NZ)
  2. Martyn Nash
    (University of Auckland, NZ)
  3. Vicky Wang
    (University of Auckland, NZ)

 

Silver Sponsors

Elite Sponsors