Multifidelity-CMA: a multifidelity approach for efficient personalisation of 3D cardiac electromechanical models

Molléro, Roch; Pennec, Xavier; Delingette, Hervé; Garny, Alan; Ayache, Nicholas; Sermesant, Maxime

doi:10.1007/s10237-017-0960-0

Cited by 21 publications

(14 citation statements)

References 30 publications

(49 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The cardiac model we use is a fast “0D” model, which we introduced in Molléro et al It is a reduced version of our in‐house 3D electromechanical model, based on the implementation of the Bestel‐Clement‐Sorine (BCS) model by Marchesseau et al() in SOFA . As described in Molléro et al, both the 3D and 0D models share the same mechanical and haemodynamic equations, but simplifying assumptions (see Appendix B) is made on the geometry of the 0D model. This leads to a very fast model made of 18 equations, which can simulate 15 beats per second at a heart rate of 75 bpm.…”

Section: Results On 137 Complete Cases and Application To Parameter Smentioning

confidence: 99%

“…To tackle this and the 3D model, we believe that a possibility would be to use the personalisation in a reduced subspace of 0D model parameters (such as

ℋ^{*}

) to influence the estimation of a 3D model. Another possibility could be to investigate the use of specific 0D/3D multifidelity parameter couplings as presented in Molléro et al for multiple iterations of the IUP algorithm at once to lower the burden of repeated personalisations. We believe this could enable the use of the IUP algorithm on databases of 100 to 200 cases with the 3D model in around 4 to 5 days.…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Population‐based priors in cardiac model personalisation for consistent parameter estimation in heterogeneous databases

Molléro

Pennec

Delingette

et al. 2018

Numer Methods Biomed Eng

Self Cite

View full text Add to dashboard Cite

Personalised cardiac models are a virtual representation of the patient heart, with parameter values for which the simulation fits the available clinical measurements. Models usually have a large number of parameters while the available data for a given patient are typically limited to a small set of measurements; thus, the parameters cannot be estimated uniquely. This is a practical obstacle for clinical applications, where accurate parameter values can be important.Here, we explore an original approach based on an algorithm called Iteratively Updated Priors (IUP), in which we perform successive personalisations of a full database through maximum a posteriori (MAP) estimation, where the prior probability at an iteration is set from the distribution of personalised parameters in the database at the previous iteration. At the convergence of the algorithm, estimated parameters of the population lie on a linear subspace of reduced (and possibly sufficient) dimension in which for each case of the database, there is a (possibly unique) parameter value for which the simulation fits the measurements. We first show how this property can help the modeller select a relevant parameter subspace for personalisation. In addition, since the resulting priors in this subspace represent the population statistics in this subspace, they can be used to perform consistent parameter estimation for cases where measurements are possibly different or missing in the database, which we illustrate with the personalisation of a heterogeneous database of 811 cases. KEYWORDScardiac electromechanical modelling, parameter estimation, parameter selection, personalised modelling INTRODUCTIONPersonalised cardiac models are of increasing interest for clinical applications. [1][2][3] To that end, parameter values of a cardiac model are estimated to get a personalised simulation that reproduces the available measurements for a clinical case. Then the personalised simulations can be used for advanced analysis of pathologies. In particular, recent works have been successful in predicting haemodynamic changes in cardiac resynchronization therapy, 4 ventricular tachycardia inducibility and dynamics, 5 and in detecting and localising infarcts 6 using 3D personalised models. Int J Numer Meth Biomed Engng. 2019;35:e3158. wileyonlinelibrary.com/journal/cnm

show abstract

Section: Results On 137 Complete Cases and Application To Parameter Smentioning

confidence: 99%

“…To tackle this and the 3D model, we believe that a possibility would be to use the personalisation in a reduced subspace of 0D model parameters (such as

ℋ^{*}

Section: Resultsmentioning

confidence: 99%

Population‐based priors in cardiac model personalisation for consistent parameter estimation in heterogeneous databases

Molléro

Pennec

Delingette

et al. 2018

Numer Methods Biomed Eng

Self Cite

View full text Add to dashboard Cite

show abstract

“…CMA-ES is a gradient-free method whose relevance was demonstrated for the personalization of complex models as in cardiac electromechanical simulations [11]. Besides, this generic personalization strategy allows to remain non-specific to a given model.…”

Section: Personalization Of the Modelsmentioning

confidence: 99%

Population-Based Personalization of Geometric Models of Myocardial Infarction

Mom

Clarysse

Duchateau

2021

Functional Imaging and Modeling of the Heart

View full text Add to dashboard Cite

We propose a strategy to perform population-based personalization of a model, to overcome the limits of case-based personalization for generating virtual populations from models that include randomness. We formulate the problem as matching the synthetic and real populations by minimizing the Kullback-Leibler divergence between their distributions. As an analytical formulation of the models is complex or even impossible, the personalization is addressed by a gradient-free method: the CMA-ES algorithm, whose relevance was demonstrated for the casebased personalization of complex biomechanical cardiac models. The algorithm iteratively adapts the covariance matrix which in our problem encodes the distribution of the synthetic data. We demonstrate the feasibility of this approach on two simple geometrical models of myocardial infarction, in 2D, to better focus on the relevance of the personalization process. Our strategy is able to reproduce the distribution of 2D myocardial infarcts from the segmented late Gadolinium images of 123 subjects with acute myocardial infarction.

show abstract

“…Here, a Levenberg‐Marquardt‐based optimization uses evaluations switching between a 3D FOM, a coarsely discretized version of the 3D FOM, and a 2D surrogate model. Another multifidelity approach was used in Reference between a 3D FOM and a 0D surrogate model. An evolutionary algorithm was used in Reference using a ROM with a pre‐computed POD‐basis from a single FOM to identify four parameters of an electrophysiological cardiac model from a synthetic electrocardiogram.…”

Section: Introductionmentioning

confidence: 99%

Using parametric model order reduction for inverse analysis of large nonlinear cardiac simulations

Pfaller

Varona

Lang

et al. 2020

Numer Methods Biomed Eng

View full text Add to dashboard Cite

Predictive high-fidelity finite element simulations of human cardiac mechanics commonly require a large number of structural degrees of freedom. Additionally, these models are often coupled with lumped-parameter models of hemodynamics. High computational demands, however, slow down model calibration and therefore limit the use of cardiac simulations in clinical practice. As cardiac models rely on several patient-specific parameters, just one solution corresponding to one specific parameter set does not at all meet clinical demands. Moreover, while solving the nonlinear problem, 90% of the computation time is spent solving linear systems of equations. We propose to reduce the structural dimension of a monolithically coupled structure-Windkessel system by projection onto a lower-dimensional subspace. We obtain a good approximation of the displacement field as well as of key scalar cardiac outputs even with very few reduced degrees of freedom, while achieving considerable speedups. For subspace generation, we use proper orthogonal decomposition of displacement snapshots. Following a brief comparison of subspace interpolation methods, we demonstrate how projection-based model order reduction can be easily integrated into a gradient-based optimization.We demonstrate the performance of our method in a real-world multivariate inverse analysis scenario. Using the presented projection-based model order reduction approach can significantly speed up model personalization and could be used for many-query tasks in a clinical setting.cardiac mechanics, inverse analysis, parametric model order reduction, proper orthogonal decomposition 1 | INTRODUCTION Cardiac solid mechanics simulations consist of solving large-deformation, materially nonlinear, elastodynamic coupled boundary-value problems. There exist different approaches to incorporate blood flow into the computational model. Three-dimensional fluid-structure interaction is resolved for example in References 1 and 2. As the exact fluid dynamics of blood within the heart are, however, usually not needed, the structural model is commonly coupled to | Model order reductionThe needed huge number of degrees of freedom (DOFs) and other challenges of solving nonlinear problems make the solution of cardiac models computationally expensive and limit the models' use in clinical practice. For example, using a single node with 24 cores, a simulation of one heartbeat, which takes about 1 s in reality, takes about 1 d to compute with our high-fidelity four chamber model. 9 The potential to reduce computation time motivates the use of reduced order models (ROMs). In this work, we solely consider model order reduction (MOR) of time-dependent parametric problems. In the following, different strategies in reduced order modeling are reviewed.An important category of cardiac ROMs is made up by simplified modeling. For these models, the same system of differential equations as for the full order model (FOM) is solved, but on a simplified analytical geometry. The displacements are commonly param...

show abstract

Multifidelity-CMA: a multifidelity approach for efficient personalisation of 3D cardiac electromechanical models

Cited by 21 publications

References 30 publications

Population‐based priors in cardiac model personalisation for consistent parameter estimation in heterogeneous databases

Population‐based priors in cardiac model personalisation for consistent parameter estimation in heterogeneous databases

Population-Based Personalization of Geometric Models of Myocardial Infarction

Using parametric model order reduction for inverse analysis of large nonlinear cardiac simulations

Contact Info

Product

Resources

About