Anne-Sophie Bonnet

Development and mechanical characterization of porous titanium bone substitutes

Commercially Pure Porous Titanium (CPPTi) can be used for surgical implants to avoid the stress shielding effect due to the mismatch between the mechanical properties of titanium and bone. Most researchers in this area deal with randomly distributed pores or simple architectures in titanium alloys. The control of porosity, pore size and distribution is necessary to obtain implants with mechanical properties close to those of bone and to ensure their osseointegration. The aim of the present work was therefore to develop and characterize such a specific porous structure. First of all, the properties of titanium made by Selective Laser Melting (SLM) were characterized through experimental testing on bulk specimens. An elementary pattern of the porous structure was then designed to mimic the orthotropic properties of the human bone following several mechanical and geometrical criteria. Finite Element Analysis (FEA) was used to optimize the pattern. A porosity of 53% and pore sizes in the range of 860 to 1500 μm were finally adopted. Tensile tests on porous samples were then carried out to validate the properties obtained numerically and identify the failure modes of the samples. Finally, FE elastoplastic analyses were performed on the porous samples in order to propose a failure criterion for the design of porous substitutes.

Biomechanical study of mandible bone supporting a four-implant retained bridge Finite element analysis of the influence of bone anisotropy and foodstuff position

PURPOSEnThe aim of this work was to study the biomechanical behavior of a All-on-four implant-supported prosthesis through a finite element analysis using either isotropic or anisotropic properties of bone. The influence of foodstuff position during mastication was also analysed.nnnMATERIALS AND METHODSnA three-dimensional finite element model of a mandible with a prosthesis supported by four implants was developed. The geometry of the edentulous mandible and prosthesis was generated from computed tomography. Four MKIII implants (two vertical and two tilted) were modeled. The bone elastic properties used in the anisotropic simulations were orthotropic. The comparison of isotropic and anisotropic models was carried out in the loading condition of mastication with a foodstuff positioned on molar. Three distinct configurations, corresponding to three foodstuff positions, were then studied. MSC/Marc code was used to perform all computations.nnnRESULTSnSignificant differences in stress, strain, and strain energy densities were found in the comparison of isotropic and orthotropic models. Molar position was revealed to be the most critical one, from a stress and strain level point of view, for implants and framework and consequently for peri-implant bone. It was also observed that implant tilting leads to high stress concentrations in bone for the All-on-four concept.nnnCONCLUSIONnThis study showed that the anisotropic behaviour of bone cannot be neglected in the numerical simulations. The actual design of the prosthesis is not optimal concerning the capacity of all metallic parts to support loads. Finally, it was demonstrated that the tilting of implants induced a high stress level at bone-implant interface.

Fatigue behavior of thin-walled grade 2 titanium samples processed by selective laser melting. Application to life prediction of porous titanium implants.

Because of its biocompatibility and high mechanical properties, the commercially pure grade 2 titanium (CPG2Ti) is largely used for fabrication of patient specific implants or hard tissue substitutes with complex shape. To avoid the stress-shielding and help their colonization by bone, prostheses with a controlled porosity are designed. The selective laser melting (SLM) is well adapted to manufacture such geometrically complicated structures constituted by struts with rough surfaces and relatively small diameters. Few studies were dedicated to characterize the fatigue properties of SLM processed samples and bulk parts. They followed conventional or standard protocols. The fatigue behavior of standard samples is very different from the one of porous raw structures. In this study, the SLM made as built (AB) and heat treated (HT) tubular samples were tested in fatigue. Wöhler curves were determined in both cases. The obtained endurance limits were equal to σD(AB)=74.5MPa and σD(HT)=65.7MPa, respectively. The heat treatment worsened the endurance limit by relaxation of negative residual stresses measured on the external surface of the samples. Modified Goodman diagram was established for raw specimens. Porous samples, based on the pattern developed by Barbas et al. (2012), were manufactured by SLM. Fatigue tests and finite element simulations performed on these samples enabled the determination of a simple rule of fatigue assessment. The method based on the stress gradient appeared as the best approach to take into account the notch influence on the fatigue life of CPG2Ti structures with a controlled porosity. The direction dependent apparent fatigue strength was found. A criterion based on the effective, or global, nominal stress was proposed taking into account the anisotropy of the porous structures. Thanks to this criterion, the usual calculation methods can be used to design bone substitutes, without a precise modelling of their internal fine porosity.

Semi-automated algorithm for cortical and trabecular bone separation from CT scans

One of the most challenging problems in modern medical imaging techniques is tissue recognition based on image segmentation procedure. The image segmentation may be described as a process of assigning a type of tissue to each pixel of the medical image. Segmentation belongs to an illposed problem class. Usually, its solution is not unique. Even results of segmentation performed manually by two specialists may slightly differ. The main difficulty in bone tissue recognition is to distinguish between cortical and trabecular parts of bone. It is relatively easy to recognise bone among other tissues, especially in computed tomography images. However, a problem arises when two types of bone have to be discriminated. The method proposed in this work combines both medical expert knowledge and skills with genetic algorithm (GA) procedure. The aim is to achieve, in a semi-automatic process saving human expert time, the best possible image segmentation. Our work follows the idea proposed by Bosco (2001) and Cagnoni et al. (1999). A two-stage procedure with learning and appropriate segmentation phases was developed. The GA approach was applied only for the first phase. In contrast to the concepts mentioned by authors, our subjects of evaluation are image-processing filters instead of segmentation curves.

Subject-specific numerical estimation of the temporomandibular joint reaction force during mouth opening and closing movements

The temporomandibular joint (TMJ) is composed of the mandibular condyle, the temporal fossa and an articular disc in between. The TMJ can be seriously damaged as a result of bone or dental disorders. To carry out joint reconstruction and design prostheses, there is a need to better understand the TMJ 3D kinematics for estimating the joint reactions and disc contribution. In the literature, mandibular movements were used to describe the TMJ kinematics but the studies were carried out on small groups (,30) (Leader et al. 2003; Ferrario et al. 2005; Coutant et al. 2008; Tuijt et al. 2010). Besides, coupled studies of TMJ kinematics and dynamics are rather rare. Finally, no comparison between healthy and pathological subjects was done. This study aims to analyse TMJ kinematics and dynamics for large groups of asymptomatic and toothless subjects.

DETERMINATION OF TIME-DEPENDENT FORCES IN HUMAN MANDIBLE DISTRACTION OSTEOGENESIS

Distraction Osteogenesis (DO) is often used for the treatment of traumas with important bone defects as ballistic wounds. External bone transport devices are thus preferred, mainly in reason of the forces encountered [Labbe, 2005]. Bone callus participates in these resisting forces and its mechanical response strongly depends on time [Richards, 1999]. Mandibular DO forces have already been studied but without considerations about immediate timedependency [Robinson, 2001]. In this context, the aim of this study was to determine the time-dependent forces supported by a distraction device (DEOS, OBL, France), see figure 1. From these measurements, the time-dependent force-displacement behaviour of bone regenerate during the early phases of human mandible DO has been evaluated.

Finite element analysis of a mandibular implant-retained overdenture

In spite of the recent efforts concerning prevention and treatment of dental diseases, total edentulism remains an important world health problem, even in industrialized countries. Different solutions to mandibular total edentulism are available from the classical removable denture to the implant supported prostheses. A possible solution is the mandibular implant-retained overdenture. Biomechanical works about this subject are longitudinal clinical studies and are limited compared with finite element analysis, in particular because of ethical reasons. Finite Element Analysis is known as an efficient method to study stresses in dental biomechanics (Castaño et al. 2002 and Cruz et al. 2003). Some attempts have been made to carry out a three-dimensional modelling of a mandible equipped with a mandibular implant-retained overdenture. However, they remain too rough from a mechanical point of view. A rigorous consideration of boundary conditions is essential to obtain a reliable model. Consequently, the aim of our study is to elaborate such a complete finite element model. Information are for example obtained about stresses and displacements acting on bone and on the overdenture retention system.

Influence of bone anisotropy in the case of implant-supported prostheses using a “All-On-Four” base

Implant-supported prostheses offer advantageous solutions for total edentulous mandibles. However, strong stresses can be expected in bones because of the implants presence. Finite element simulations are frenquently used to reproduce the behaviour of a mandibule supporting such prosthesis. It is necessary, in all cases, to define properly geometry, boundary conditions and material properties to guarantee results’ accuracy. Because of their high level of heterogeneity and anisotropy, bones behaviour is very difficult to model. The effect of bones anisotropy in the premolar region was analysed by O’Mahonny et al. (2001). In this work, we propose to study this effect in the mandible symphyseal region, for the case of an implantsupported prosthesis using an “All-On-Four” base. The traditional length of implant-supported prostheses is limited by the position of posterior implants that cannot interact with the mental nerve. Consequently, the number of restored teeth is generally limited to 10. To increase this number Krekmanov et al. (2000) studied the possibility to tilt posterior implants.