The objective of Computational Physiology (CP) is to provide models of the major functions of the human body and numerical methods to simulate them. The main applications are in medicine and biology, where CP can be used for instance to better understand the basic processes leading to the apparition of a pathology, to model its probable evolution and to plan, simulate, and monitor its therapy.
Quite advanced models have already been proposed to study at the molecular, cellular and organic level a number of physiological systems. While these models and new ones need to be developed, reﬁned or validated, a grand challenge that we want to address in this project is the automatic adaptation of the model to a given patient by confronting the model with the available biomedical
images and signals and possibly also from some additional information (e.g. genetic). Building such patientspeciﬁc models is an ambitious goal which requires the choice or construction of models with a complexity adapted to the resolution of the accessible measurements and the development of new data assimilation methods coping with massive numbers of measurements and unknowns.
There is a hierarchy of modeling levels for CP models of the human body:
- the ﬁrst level is mainly geometrical, and addresses the construction of a digital description of the anatomy, essentially acquired from medical imagery;
- the second level is physical, involving mainly the biomechanical modeling of various tissues, organs, vessels, muscles or bone structures;
- the third level is physiological, involving a modeling of the functions of the major biological systems (e.g. cardiovascular, respiratory, digestive, central or peripheral nervous, muscular, reproductive, hormonal, etc.) or some pathological metabolism (e.g. evolution of cancerous or inﬂammatory lesions, formation of vessel stenoses, etc.);
- a fourth level would be cognitive, modeling the higher functions of the human brain.
These different levels of modeling are closely related to each other, and several physiological systems may interact together (e.g. the cardiopulmonary interaction). The choice of the resolution at which each level is described is important, and may vary from microscopic to macroscopic, ideally through multiscale descriptions.
Building this complete hierarchy of models is necessary to evolve from a Visible Human project (essentially ﬁrst level of modeling) to a much more ambitious Physiological Human project. We will not address all the issues raised by this ambitious project, but instead focus on topics detailed below. Among them, our objective is to identify some common methods for the resolution of the large inverse problems raised by the coupling of physiological models to biological images for the construction of patient-speciﬁc models (e.g. speciﬁc variational or sequential methods (EKF), dedicated particle ﬁlters, etc.). We also plan to develop a speciﬁc expertise on the extraction of geometrical meshes from medical images for their further use in simulation procedures. Finally, computational models can be used for speciﬁc image analysis problems (e.g. segmentation, registration, tracking, etc.). Application domains include
1. Surgery Simulation,
2. Cardiac Imaging,
3. Brain tumors, neo-angiogenesis,…
In-silico tumor growth
We propose to model and simulate the growth of glioblastomas, the most aggressive of the glial tumors. The proposed simulation is based on a model coupling the invasion of the glioblastoma and its mechanical interaction with the invaded structures. This model uses a reaction-diffusion equation for the tumor expansion characterization and the usual continuum mechanics laws for the brain parenchyma behavior. In addition, we propose a new equation to couple these two equations to take into account the mechanical influence of the tumor cells on the invaded tissues.
Our model relies upon an anatomical atlas including cerebral structures having a distinct response to the tumor aggression. In addition, we included in this atlas the information from the Diffusion Tensor Images (DTI) to model the tumor preferential growth in the white fibers direction. Finally, the tumor growth model is used to simulate a virtual GBM grow into a patient brain. This in-silico growth is compared to the real GBM growth observed with a second patient MRI taken 6 months later.More details can be found here
Related publications can be found here
Participants: Erin Stretton [Correspondant], Nicholas Ayache, Hervé Delingette, Bjoern Menze, Ender Konukoglu, Ezequiel Geremia, Emmanuel Mandonnet.
This work was funded by Care4me.
glioma, tumor modeling, reaction diffusion equation, traveling wave equation, DTI, MRI
• Performing a sensitivity analysis of long time series of multi-modal data.
• Improving our current models and their inputs, including creating a white matter and brain mask template with the help of a neurosurgeon and integrating the Powell bound constrained optimization into the minimization routine.
• Performed comparison of using a patient DTI, an atlas DTI or no DTI at all showing the difference in accuracies in the simulation results since the patient DTI is not always available in a clinical setting or is of poor quality. We found that using an atlas DTI produced only slightly less accurate results than using a patient DTI, where as using no DTI at all did not produce accurate results.
Synthetic Echocardiographic Image Sequences for Cardiac Inverse Electro-Kinematic Learning
Participants: Adityo Prakosa [Correspondant], Maxime Sermesant, Hervé Delingette, Eric Saloux [CHU Caen], Pascal Allain [Philips Healthcare], Pascal Cathier [Philips Healthcare], Patrick Etyngier [Philips Healthcare], Nicolas Villain [Philips Healthcare], Nicholas Ayache.
This work is done in collaboration with Medisys, Philips Healthcare Suresnes, France, and Cardiology Department of CHU Caen, France.
synthetic 3D echocardiography, non-rigid registration, cardiac electromechanical model, inverse electro-kinematic learning, Least-Square Support Vector Machine
• A database of 120 synthetic 3D echocardiography (US) sequences is created based on a cardiac electromechanical model.
• Kinematic descriptors are extracted from the displacement field estimated from the synthetic 3D US sequence using the iLogDemons non-rigid registration method.
• Cardiac inverse electro-kinematic learning is done by using the database of synthetic 3D US sequences in order to estimate the cardiac depolarization times for the given kinematic descriptors. First evaluation on two clinical sequences from patients with Left Bundle Branch Block shows en- couraging results.
Prediction of patient-specific Ventricular Tachycardia for radio-frequency ablation therapy planning
Participants: Jatin Relan [Correspondant], Maxime Sermesant, Hervé Delingette, Nicholas Ayache. This work is funded by the FP7 European Project euHeart.
Cardiac Electrophysiology Modelling, Reaction-Diffusion Models, Parameter estimation, Arrhythmia prediction
In this work, we build a patient-specific cardiac electrophysiology (EP) model derived from hybrid XMR imaging and non-contact electro-anatomical mapping procedure on a patient with heart failure. The model is then used to predict patient-specific arrhythmias, such as induced ischemic Ventricular Tachycardia (VT) and leads in generation and evaluation of patient-specific VT circuits, with critical exit points for Radio Frequency (RF) ablation. These predictions are now validated with some clinical cases, with electrophysiology mapping of induced VT in patients undergoing the clinical VT-Stim procedures.
Participants: Hugo Talbot [Correspondant], Federico Spadoni, Maxime Sermesant, Hervé Delingette, Stephane Cotin.
This work is performed in the context of the euHeart project and the PhD of Hugo Talbot in collaboration with the Shacra (INRIA, Lille Nord Europe) team.
Electrophysiology, simulation, real-time, cardiac arrhythmia, ablation procedure.
• This work aims at simulating in real-time the endovascular procedure of radiofrequency ablation of the left ventricle for patient suffering from Ventricular Tachycardia.
• Fast simulation of electrophysiology has been reached with a Eikonal model.
• Use the SOFA platform for simulating endovascular navigation and cardiac electrophysiology.
Participants: Stéphanie Marchesseau [Correspondant], Ken C.L. Wong, Hervé Delingette, Maxime Serme- sant, Nicholas Ayache.
This work has been performed in the context of the euHeart european project.
Cardiac simulation, sensitivity analysis
• We improved the existing electromechanical model of the heart to include mechanical non linearity, viscosity and strain rate dependent contractility. It was implemented in SOFA with a new four valves model to deal with the cardiac phases and enforce isovolumetric constraint.
• We have obtained first personalization of cardiac mechanics from 3 cine-MRI cases using a variational approach (adjoint method).