Disease-Based Models of Cardiac Electromechanics
6. Conclusion
Disease-based models of cardiac electromechanics have been presented. The models, based on the method previously developed at INRIA (Sermesant et al., 2006), has been tailored to values observed in children. The geometry has been adapted before-hand in order to get consistent a volume ratio between the right and left cavities, and the normal heart has been simulated to calibrate the electromechanical parameters and to get simulated clinical parameters consistent with the values measured in healthy children.
Once the calibration has been done, three simulations of right-ventricle overload (RVO), dilated cardiomyopathy (DCM) and hypertrophic cardiomyopathy (HCM) have been performed. In these experiments, only the most relevant clinical parameters have been used to adapt the standard heart model to the pathology. In RVO, the geometry of the right ventricle and the pressure of the right atrium have been changed. In DCM, both cardiac cavities have been dilated and the contraction have been reduced to simulate the impaired systolic function. Finally, in HCM, left-ventricular myocardium has been thickened and the contraction and relaxation rate have been reduced.
The obtained results are promising. In each case, the main clinical observations are recovered: it is thus possible to reliability simulate pathologies.
Further work is however required to improve the results. In these experiments, some parameters, like the maximum contraction for example, have been adjusted so as to get the expected results. Of course, the changes have been done accordingly with respect to clinical surveys but the chosen values may not be consistent with the actual measurements. Nevertheless, this approach is mandatory to design an accurate and reliable disease-based model: it will allow to devise new clinical measurements that would better quantify well-known parameters (the contractility of the myocardium for instance) or indeed find hidden relationships between various measurements.
The next stages in the development of such models will then consist in improving the simulated heart geometry, implementing focal variabilities to simulate localised pathological features, and enhance the modelling of the boundary conditions.
