Behzad Shariat

Simulation of lung behaviour with finite elements: influence of bio-mechanical parameters

Motivated by medical needs, we propose to simulate lung deformation and motion during respiration to track tumours. This paper presents a model of lung behaviour based on a continuous media mechanics model and solved with a finite element method. The result is a simulation of a normal breathing, matching with patient customised data. Moreover, we carried out numerical experiments to evaluate our algorithms and to measure the influence and the relevance of mechanical parameters.

Biomechanical simulation of the fetal descent without imposed theoretical trajectory

The medical training concerning childbirth for young obstetricians involves performing real deliveries, under supervision. This medical procedure becomes more complicated when instrumented deliveries requiring the use of forceps or suction cups become necessary. For this reason, the use of a versatile, configurable childbirth simulator, taking into account different anatomical and pathological cases, would provide an important benefit in the training of obstetricians, and improve medical procedures. The production of this type of simulator should be generally based on a computerized birth simulation, enabling the computation of the reproductive organs deformation of the parturient woman and fetal interactions as well as the calculation of efforts produced during the second stage of labor. In this paper, we present a geometrical and biomechanical modeling of the main parturients organs involved in the birth process, interacting with the fetus. Instead of searching for absolute precision, we search to find a good compromise between accuracy and model complexity. At this stage, to verify the correctness of our hypothesis, we use finite element analysis because of its reliability, precision and stability. Moreover, our study improves the previous work carried out on childbirth simulators because: (a) our childbirth model takes into account all the major organs involved in birth process, thus potentially enabling different childbirth scenarios; (b) fetal head is not treated as a rigid body and its motion is computed by taking into account realistic boundary conditions, i.e. we do not impose a pre-computed fetal trajectory; (c) we take into account the cyclic uterine contractions as well as voluntary efforts produced by the muscles of the abdomen; (d) a slight pressure is added inside the abdomen, representing the residual muscle tone. The next stage of our work will concern the optimization of our numerical resolution approach to obtain interactive time simulation, enabling it to be coupled to our haptic device.

Deformable object reconstruction with particle systems

Abstract In this paper, we describe a general method for reconstruction and animation of volumetric deformable objects. First, the initial data, defined by planar contours, are fitted with a closed periodic parametric surface. Then, this rigid model is used as a closed boundary, which is filled in with particles. For this, new particles are progressively generated within the boundary. When a new particle is introduced, there will be an interaction between this particle and all the existing particles. They will also collide the boundaries. Then we let the system evolve under physically-based forces, until the particles reach an equilibrium state. We have defined some simple rules for the new particle generation. In the scope of a medical application, we want to simulate the motion and the form alteration of the internal anatomical organs. Hence we have introduced the necessary tools to handle the dynamic and deformable behavior of cancerous organs.

Semantic interoperability of knowledge in feature-based CAD models

Dans un environnement collaboratif de developpement de produit, plusieurs acteurs, ayant differents points de vue et intervenant dans plusieurs phases du cycle de vie de produit, doivent communiquer et echanger des connaissances entre eux. Ces connaissances, existant sous differents formats heterogenes, incluent potentiellement plusieurs concepts tels que l’historique de conception, la structure du produit, les features, les parametres, les contraintes, et d’autres informations sur le produit. Les exigences industrielles de reduction du temps et du cout de production necessitent l’amelioration de l’interoperabilite semantique entre les differents processus de developpement afin de surmonter ces problemes d’heterogeneite tant au niveau syntaxique, structurel, que semantique. Dans le domaine de la CAO, la plupart des methodes existantes pour l’echange de donnees d’un modele de produit sont, effectivement, basees sur le transfert des donnees geometriques. Cependant, ces donnees ne sont pas suffisantes pour saisir la semantique des donnees, telle que l’intention de conception, ainsi que l’edition des modeles apres leur echange. De ce fait, nous nous sommes interesses a l’echange des modeles « intelligents », autrement dit, definis en termes d’historique de construction, de fonctions intelligentes de conception appelees features, y compris les parametres et les contraintes. L’objectif de notre these est de concevoir des methodes permettant d’ameliorer l’interoperabilite semantique des systemes CAO moyennant les technologies du Web Semantique comme les ontologies OWL DL et le langage des regles SWRL. Nous avons donc elabore une approche d’echange basee sur une ontologie commune de features de conception, que nous avons appelee CDFO « Common Design Features Ontology », servant d’intermediaire entre les differents systemes CAO. Cette approche s’appuie principalement sur deux grandes etapes. La premiere etape consiste en une homogeneisation des formats de representation des modeles CAO vers un format pivot, en l’occurrence OWL DL. Cette homogeneisation sert a traiter les heterogeneites syntaxiques entre les formats des modeles. La deuxieme etape consiste a definir des regles permettant la mise en correspondance semantique entre les ontologies d’application de CAO et notre ontologie commune. Cette methode de mise en correspondance se base principalement, d’une part, sur la definition explicite des axiomes et des regles de correspondance permettant l’alignement des entites de differentes ontologies, et d’autre part sur la reconnaissance automatique des correspondances semantiques supplementaires a l’aide des capacites de raisonnement fournies par les moteurs d’inference bases sur les logiques de description. Enfin, notre methode de mise en correspondance est enrichie par le developpement d’une methode de calcul de similarite semantique appropriee pour le langage OWL DL, qui repose principalement sur les composants des entites en question tels que leur description et leur contexte.

Lung mesh generation to simulate breathing motion with a finite element method

Numerical modelling of lung behaviour during the respiration cycle is a difficult challenge due to its complex geometry and surrounding environment constraints. This paper presents an approach to simulate a patients lung motion during inhaling and exhaling based on a continuous media mechanics model and solved with a finite element method. One of the key problems is an adequate lung mesh generation, which is specifically developed in this paper.

Breathing Thorax Simulation based on Pleura Physiology and Rib Kinematics

To monitor a lung mechanical model and then predict tumour motion we proposed a approach based on the pleura physiology. By comparing the predictions to landmarks set by medical experts, we observed better results with regards to the one obtained with approaches found in the literature. Beside, we focus on the rib cage kinematics, which play a significant role in the pleura outer-surface motion and therefore in the lung motion. We proposed a kinematic model of the rib cage based on the finite helical axis method and we show out interesting results.

A biomechanical model of the female reproductive system and the fetus for the realization of a childbirth virtual simulator

Our main work consists in modeling of the female pelvis and uterus, as well as the human fetus. The goal of this work is to recover the different forces generated during the delivery. These forces will be input to the haptic obstetric training tool BirthSim which has already been developed by the Ampère Laboratory at the INSA of Lyon. This modeling process will permit us to develop a new training device to take into account different anatomies and different types of delivery. In this paper, we will firstly show the different existing haptic and virtual simulators in the obstetric world with their advantages and drawbacks. After, we will present our approach based on a biomechanical modeling of concerned organs. To obtain interactive time performance, we proceed by the simplification of the organs anatomy. Then, we present some results showing that FEM analysis can be used to model forces during childbirth. In the future, we plan to use this work to more accurately control a childbirth simulator.

A Chest Wall Model Based on Rib Kinematics

The success of radiotherapy treatment could be compromised by motion. Lung tumours are particularly concerned by this problem because their positions are subject to breathing motion. To reduce the uncertainty on the position of pulmonary tumours during breathing cycle, we propose to develop a complete thoracic biomechanical model. This model will be monitored through the measurement of external parameters (thorax outer-surface motion, air flow...) and should predict in real-time the location of lung tumour. In this paper, we expose a biomechanical model of the lung environment, based on anatomical and physiological knowledge. The model includes the skin, the ribs, the pleura and the soft tissue between the skin and the ribcage. Motions and deformations are computed with the Finite Element Method. The ribcage direct kinematics model, permits to compute the skin position from the ribs motion. Conversely, the inverse kinematics provides rib motion and consequently lung motion. It can be computed from the outer-surface motion.With regards to available clinical data the results are promising. In particular, the average error is lower than the resolution of the CT-scan images used as input data.

Towards accurate tumour tracking in lungs

Motivated by radiotherapy and hadrontherapy improvement, we consider in a first step the potential of a simple elastic mechanical modelling to simulate lung deformations and motions during respiration, towards tumour tracking. Two approaches are presented: one is the finite-element based method and the other is the mass-spring system. For these approaches, we suggest a personalisation based on the measurement of physical and geometrical data for each patient.

Towards an Intelligent CAD Models Sharing Based on Semantic Web Technologies

Interoperability among CAD systems is a well known problem in collaborative product design and development. Nowadays, existing solutions and standards for product data integration are restricted to the process of geometric data. As a result, the model can hardly be modified, and the original intent of the designer may be misunderstood. Hence, design intent such as construction history, features, parameters and constraints should be consistently maintained while processing semantically modeling terms both by design collaborators and intelligent systems. In this paper, we investigate the use of Semantic Web technologies for the development of a common design features ontology, sharable for collaborative design. In our approach, we define the neutral format as an ontology using OWL (Web Ontology Language), and more specifically its sublanguage OWL DL based on Description Logics, which is then enriched from the logical data model with axioms and rules. Rules have been defined using SWRL (Semantic Web Rule Language) for enriching our ontology expressivity such as for handling composed properties.