Dynamic contrast enhanced MR imaging (DCE-MRI) has commonly been used to investigate disruptions in microvascular capillary permeability in pathologies such as tumours, and in brain diseases such as multiple sclerosis. This imaging technique involves intravenous injection of a contrast agent, which can modulate MR signal contrast, while frequently acquiring images (i.e. every few seconds) as the agent perfuses through the tissue of interest. Microvascular permeability, and other parameters such as blood volume and flow (perfusion) can be quantified through application of a pharmacokinetic model on the data acquired from the MRI scan. The model requires input from both the biological (e.g. pharmacokinetic rate constants) as well as physical (i.e. scanner settings) parameters. As there are a great many variables and different biophysical models (e.g. high blood flow, high permeability tissues, etc.) there needs to be considerations made for situations where the permeability may be only slightly different from normal. In the brain the blood-brain barrier (BBB) is a highly selective barrier that restricts most bulk diffusion/permeability of solutes. Changes in BBB permeability is likely only subtle in diseases such as depression or bipolar disorder, especially when compared against hypervascular-hypermemeable cancers that are void of a BBB altogether. The problem is however, to decide which model of BBB permeability is best suited for differentiating subtle changes. Thus the intention of this project was to investigate multiple pharmacokinetic models for the tracking of MRI contrast agent in regions of the brain with an intact BBB. In the brain, where there is strict regulation of molecules passing through the microvasculature into the extracellular space, and where more subtle disruptions might be of interest, different assumptions may be necessary. Four models were investigated: the Tofts model, the modified Tofts model, the two-compartment exchange model, and the uptake model. Scans of eight healthy subjects were analyzed, and permeability was quantified using each model. The accuracy of each model, quantified by the R\textsuperscript{2} value, were compared. Analysis found that the Tofts model performed significantly worse than the modified Tofts and Uptake models when fitting regions of the brain with a blood-brain barrier, with a p-value of 0.006. The analysis did not reveal any significant difference between the modified Tofts, Uptake or 2CX models, although perhaps it was obscured due to the limited number of data points. Further investigation is needed to determine any differences between the three top-performing models. / Thesis / Master of Applied Science (MASc)

Understanding the processes that govern pre-systemic drug absorption and elimination is of high importance in pharmaceutical research and development, and clinical pharmacotherapy, as the oral route remains the most frequently used route of drug administration. The emergence of systems pharmacology has enabled the utilisation of in silico physiologically-based pharmacokinetic (PBPK) modelling and simulation (M&S) coupled to in vitro-in vivo extrapolation in order to perform extrapolation and exploratory M&S in special populations and scenarios were concerns regarding alterations in oral drug exposure may arise, such as following gastrointestinal (GI) surgery or metabolic drug-drug interactions (DDIs).Due to the multi-factorial physiological implications of bariatric surgery, resulting in the partial resection of the GI tract, the inability to rationalise and predict trends in oral drug bioavailability (Foral) following surgery present considerable pharmacotherapeutical challenges. PBPK M&S is a highly implemented approach for the prediction of DDIs. Reoccurring issues have emerged with regards to predictions of the magnitude of mechanism-based inhibition (MBI) where overestimations of DDIs have repeatedly been reported for drugs exhibiting high intestinal extraction. The aim of this thesis was to explore the interplay between oral drug absorption and metabolism occurring in the GI tract through the exploration of the impact of bariatric surgery on oral drug exposure and by theoretically examining the nesting and hierarchy of enterocyte and enzyme turnover and its impact on MBIs in the small intestine. This would be carried out by utilising a systems pharmacology PBPK M&S approach under a general model development framework of identification and characterisation of critical intrinsic factors and parameters, model implementation and validation. Developed post bariatric surgery PBPK models allow a framework to theoretically explore physiological mechanisms associated with altered oral drug exposure pre to post surgery, which could be assigned to the interplay between dissolution, absorption and gut-wall metabolism, where dissolution and formulation properties emerged as the perhaps most important parameters in predicting the drug disposition following surgery. Model validation identified missing critical factors that are essential for additional model refinement. Developed post bariatric surgery PBPK models have the potential of aiding clinical pharmacotherapy and decision-making following surgery. A mechanistic PBPK model was developed to describe the hierarchical dependency of enzyme and enterocyte turnover in the small intestine. Predicted enzyme recovery using the nested enzyme-within-enterocyte turnover model may potentially account for reported overpredictions of mechanism-based inhibition. Developed models in this thesis showcase the advantage of PBPK M&S in the extrapolation of oral drug exposure to special population and the potential of a PBPK approach in understanding underlying the underlying mechanism governing Foral and additionally highlight the need for generation of interdisciplinary data to support model development.

Produced since the 1950’s, per- and polyfluoroalkyl (PFASs) substances are persistent, bioaccumulative and toxic compounds that are ubiquitous in the environment. Being proteinophilic with a tendency to partition to protein-rich tissues, PFASs have been found in human serum worldwide and in wildlife with a predominance of long-chain perfluoroalkyl carboxilic acids (C7-C14 PFCAs) and perfluoroalkyl sulfonic acids (C6-C9 PFSAs). Due to rising concern regarding their hazardous properties, several regulatory actions and voluntary industrial phase-outs have been conducted since early 2000s, shifting the production towards other fluorinated alternatives. This thesis explores the human exposure to long-chain PFASs and their alternatives using different modelling methods and aims to 1) link comprehensively the past and current industrial production with the human body burden and 2) assess the potential hazardous properties of legacy PFASs replacements, on which information is very limited. In Paper I, the historical daily intakes in Australia and USA were reconstructed from cross-sectional biomonitoring data of perfluorooctanesulfonic acid (PFOS), perfluorooctanoic acid (PFOA) andperfluorohexanesulfonic acid (PFHxS). The results indicate that humans experienced similar exposure levels and trends to PFOS and PFOA in both regions, suggesting a common historical exposure possibly dominated by consumer products. The model could not be fitted to PFHxS concentration in serum. In Paper II, the relative contribution of indirect (i.e. subsequent metabolism of precursors into legacy PFASs) versus direct exposure was evaluated on occupationally exposed ski wax technicians. The indirect exposure contributed by up to 45% to the total body burden of PFOA. In Paper III, the physicochemical properties, the persistence and the long-range transport of fluorinated alternatives were predicted using different in silico tools. Findings suggest that fluorinated alternatives are likely similar to their predecessors, in terms of physicochemical properties and environmental fate. Finally, Paper IV compares the toxic potency of PFOS, PFOA and their alternatives as a function of external and internal dose. While alternatives are less potent than their predecessors when considering the administered dose, they become similarly potent when the assessment is based on levels in the target tissue. This thesis demonstrates that pharmacokinetic models are effective tools to comprehensively reconnect the body burden to the exposure of phased-out chemicals. More importantly, the studies on fluorinated alternatives raise the necessity to provide more information and data on the potential hazard of these novel and emerging products. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 1: Manuscript. Paper 4: Manuscript.</p>

Pharmacokinetic and pharmacodynamic models can be applied to clinical study data using various modelling approaches depending on the aim of the analysis. In population pharmacokinetics for instance, simple compartmental models can be employed to describe concentration-time data, identify prognostic factors and interpolate within well-defined experimental conditions. The first objective of this thesis was to illustrate such a ‘semi-mechanistic’ pharmacokinetic modelling approach using mavoglurant as an example of a compound under clinical development. In particular, methods to accurately characterise complex oral pharmacokinetic profiles and evaluate the impact of absorption factors were investigated. When the purpose of the model-based analysis is to further extrapolate beyond the experimental conditions in order to guide the design of subsequent clinical trials, physiologically-based pharmacokinetic (PBPK) models are more valuable as they incorporate information not only on the drug but also on the system, i.e. on mammillary anatomy and physiology. The combination of such mechanistic models with statistical modelling techniques in order to analysis clinical data has been widely applied in toxicokinetics but has only recently received increasing interest in pharmacokinetics. This is probably because, due to the higher complexity of PBPK models compared to conventional pharmacokinetic models, additional efforts are required for adequate population data analysis. Hence, the second objective of this thesis was to explore methods to allow the application of PBPK models to clinical study data, such as the Bayesian approach or model order reduction techniques, and propose a general mechanistic modelling workflow for population data analysis. In pharmacodynamics, mechanistic modelling of clinical data is even less common than in pharmacokinetics. This is probably because our understanding of the interaction between therapeutic drugs and biological processes is limited and also because the types of data to analyse are often more complex than pharmacokinetic data. In oncology for instance, the most widely used clinical endpoint to evaluate the benefit of an experimental treatment is survival of patients. Survival data are typically censored due to logistic constraints associated with patient follow-up. Hence, the analysis of survival data requires specific statistical techniques. Longitudinal tumour size data have been increasingly used to assess treatment response for solid tumours. In particular, the survival prognostic value of measures derived from such data has been recently evaluated for various types of cancer although not for pancreatic cancer. The last objective of this thesis was therefore to investigate different modelling approaches to analyse survival data of pancreatic cancer patients treated with gemcitabine, and compare tumour burden measures with other patient clinical characteristics and established risk factors, in terms of predictive value for survival.

Abnormal vascular structure has been identified as one of the major characteristics of tumours. In this thesis, we carry out quantitative analysis on different tumour vascular structures and research the relationship between vascular structure and its transportation efficiency. We first study segmentation methods to extract the binary vessel representations from microscope images. We found that local phase-hysteresis thresholding is able to segment vessel objects from noisy microscope images. We also study methods to extract the centre lines of segmented vessel objects, a process termed as skeletonization. We modified the conventional thinning method to regularize the extremely asymmetrical structure found in the segmented vessel objects. We found this method is capable to produce vessel skeletons with satisfactory accuracy. We have developed a software for 3D vessel structural analysis. This software is consisted of four major parts: image segmentation, vessel skeletonization, skeleton modification and structure quantification. This software has implemented local phase-hysteresis thresholding and structure regularization-thinning methods. A GUI was introduced to enable users to alter the skeleton structures based on their subjective judgements. Radius and inter branch length quantification can be conducted based on the segmentation and skeletonization results. The accuracy of segmentation, skeletonization and quantification methods have been tested on several synthesized data sets. The change of tumour vascular structure after drug treatment was then investigated. We proposed metrics to quantify tumour vascular geometry and statistically analysed the effect of tested drugs on normalizing tumour vascular structure. finally, we developed a spatio-temporal model to simulate the delivery of oxygen and 3-18 F-fluoro-1-(2-nitro-1-imidazolyl)-2-propanol (Fmiso), which is the hypoxia tracer that gives out PET signal in an Fmiso PET scanning. This model is based on compartmental models, but also considers the spatial diffusion of oxygen and Fmiso. We validated our model on in vitro spheroid data and simulated the oxygen and Fmiso distribution on the segmented vessel images. We contend that the tumour Fmiso distribution (as observed in Fmiso PET imaging) is caused by the abnormal tumour vascular structure which is further aroused from tumour angiogenesis process. We depicted a modelling framework to research the relationships between tumour angiogenesis, vessel structure and Fmiso distribution, which is going to be the focus of our future work.

This dissertation deals with quantitative perfusion analysis of MRI contrast-enhanced image time sequences. It focuses on two so far separately used methods -- Dynamic contrast-enhanced MRI (DCE-MRI) and Dynamic susceptibility contrast MRI (DSC-MRI). The common problem of such perfusion analyses is the unreliability of perfusion parameters estimation. This penalizes usage of these unique techniques on a regular basis. The presented methods are intended to improve these drawbacks, especially the problems with quantification in DSC in case of contrast agent extravasation and instability of the deconvolution process in DCE using advanced pharmacokinetic models. There are a few approaches in literature combining DCE and DSC to estimate new parameters of the examined tissue, namely the relaxivity of the vascular and of the interstitial space. Originally, in this scheme, the 2CXM DCE model was used. Here various models for DCE analysis are tested keeping in mind the DCE-DSC combination. The ATH model was found to perform better in this setting compared to 2CXM. Finally, the ATH model was used in alternating DCE-DSC optimization algorithm and then in a truly fully simultaneous DCE-DSC. The processing was tested using simulated and in-vivo data. According to the results, the proposed simultaneous algorithm performs better in comparison with sequential DCE-DSC, unleashing full potential of perfusion analysis using MRI.

Outputs of dose–response assessments can be used as benchmarks that help to identify the need for risk management measures to reduce population health risks associated with exposure to chemicals. Various approaches can be used to facilitate the incorporation of chemical- or scenario-specific data into dose–response analyses, as a means of replacing or influencing default assumptions and extrapolations. The goal of the first part of this thesis was to examine the evolution of approaches to the incorporation of chemical- and scenario-specific data in dose–response assessments in regulatory settings, and identify contextual factors that serve as barriers and facilitators to the use of approaches. A main focus of the investigation was on physiological modelling, which is the most commonly-used category of approaches enabling extrapolations that depart from default assumptions. Evaluations of the dose–response applications of physiological modelling in the peer-reviewed scientific literature and in regulatory reports were conducted. Similarities between the scientific literature databases and regulatory reports were observed with respect to the evolution of physiological modelling in dose–response assessments, notably related to the timing, quantity, and annual frequency of publications. These similarities indicate that a factor in the low dose–response application of physiological modelling, relative to the overall production of physiological models, is an absence of data. However, variability in adoption of physiological modelling in regulatory dose–response assessments was observed among—and even within—organizations faced with the same data, indicating that other factors influence regulatory uptake of physiological modelling. Analysis of a survey indicated that factors acting as barriers or facilitators to regulatory risk assessors’ incorporation of increasingly data-informed approaches originated in both external and internal contexts. The external context was composed of the regulatory environment, domestic and international alignment, availability of external expertise, background of peer reviewers and stakeholders, availability and accessibility of software and tools, and chemical-dependent factors. The internal context was influenced by problem formulation, time and financial resources, organizational and management support, and training. A conceptual framework demonstrating how these factors impact a risk assessor’s ability to incorporate chemical- and scenario-specific data in dose–response analysis was developed, and subsequently used to provide recommendations on actions that could be taken to increase regulatory adoption of increasingly data-informed approaches. The second part of the thesis focused on the development of a knowledge translation tool designed to assist risk managers in the evaluation of dose–response analyses. The tool was focused on occupational exposure limits (OELs), and provides a guide to occupational hygienists in evaluating the relevance and reliability of individual OELs. When occupational hygienists are faced with multiple varying OELs for a chemical of interest, these evaluations can support the selection of the most appropriate OEL for a given situation. The usefulness of the tool was demonstrated for the selection of OELs for an OEL-rich compound (n-hexane), an OEL-poor compound (methamphetamine), and one additional compound (manganese). Such a tool can improve occupational hygienists’ understanding of the basis of OELs and the levels of protection afforded by each, which can contribute to more informed risk management decisions.

Un nouveau règlement (CE) n° 1901/2006 établi par le Parlement européen et le Conseil de l’UE, relatif aux médicaments à usage pédiatrique, vise à améliorer la santé et la qualité de vie des enfants en Europe, en garantissant que les nouveaux médicaments pédiatriques et les médicaments déjà commercialisés seront pleinement adaptés à leurs besoins spécifiques. Ce règlement prévoit de nouvelles obligations pour l'industrie pharmaceutique, assorties de récompenses et d'incitations. Dans ce contexte, un plan d’investigation pédiatrique a été proposé pour l’ivabradine dans plusieurs sous-groupes de la population pédiatrique dans le traitement de l’insuffisance cardiaque chronique. L’ivabradine est une molécule déjà commercialisée chez l’adulte dans la prise en charge de l’angor, et de l’insuffisance cardiaque. Un premier travail a été d’aider au design de cette étude pédiatrique : évaluer la formulation pédiatrique, aider au choix de la dose initiale à administrer chez l’enfant, choisir le protocole de prélèvements et conseiller la méthode de prélèvements. Pour évaluer la formulation pédiatrique, une étude a été conduite pour déterminer la biodisponibilité relative de la formulation pédiatrique par rapport aux comprimés utilisés chez l’adulte. Une biodisponibilité relative similaire a été retrouvée entre les deux formulations. Une approche physiologique (PBPK « Physiollogically based PharmacoKineticsmodel ») a été utilisé pour prédire la dose initiale à administrer et pour proposer un protocole de prélèvements PK. La méthode DBS (Dried blood spot) consistant à collecter à chaque temps de prélèvement une goutte de sang (au pli du coude ou au bout du doigt) a été recommandée. La première dose à administrer chez l’enfant peut être également être déterminée par des modèles de population développés chez l’adulte et adaptés à l’enfant grâce à l’allométrie et à l’ajout de fonctions de maturation. Cette approche a été comparée au PBPK dans le cas de l’ivabradine et des résultats similaires ont été obtenus. Un deuxième travail a été réalisé après que l’étude clinique ait été conduite dans la population pédiatrique. L’étude a été menée chez 116 enfants (74 enfants recevant l’ivabradine, 42 recevant le placebo) âgés de 6 mois à 18 ans et les données ont été analysées. Tout d’abord, une relation a été établie entre les concentrations d’ivabradine plasmatiques et les concentrations d’ivabradine mesurées dans le sang total. Puis, afin de décrire les concentrations d’ivabradine et de son métabolite, un modèle de population prenant en compte l’effet de l’âge et du poids a été développé. En comparant les expositions plasmatiques, une dose par kilogramme plus élevée aurait été nécessaire chez les patients les plus jeunes pour atteindre un niveau d’exposition similaire aux patients plus âgés. Enfin, il a été monté que la relation PK/PD qui avait développé chez l’adulte était conservée dans la population pédiatrique. / New legislation governing the development and authorization of medicines for use in children was introduced in the European Union (EU) in January 2007. This Regulation aims to facilitate the development and accessibility of medicinal products for use in the paediatric population, to ensure that medicinal products used to treat the paediatric population are subject to ethical research of high quality and are appropriately authorised for use in the paediatric population, and to improve the information available on the use of medicinal products in the various paediatric populations. Several rewards and incentives for the development of paediatric medicines for children are available in the European Union (EU). In compliance with the paediatric European regulation, a study will be conducted in paediatric patients with CHF with the objective to determine the efficacious and safe dose of ivabradine, a compound already marketed in adults, and to assess its efficacy and safety in children over 1 year old. A first work was to help design a paediatric study for ivabradine focusing on: the paediatric formulation evaluation, the doses to be administered, the sampling design and the sampling technique. A study was conducted in order to assess the relative bioavailability (Frel) of the paediatric formulation and a similar Frel was observed between the paediatric formulation and the adult marketed tablet. PBPK modelling was used to predict initial doses to be administered in the paediatric study and to select the most appropriate sample time collections. The dried blood spot (DBS) technique was recommended in the clinical trial in children. A secondary objective was to perform a comparison of the prediction of ivabradine pharmacokinetics (PK) in children using a physiologically-based pharmacokinetic (PBPK) approach and allometric scaling of a population pharmacokinetic (PPK) model. Simulations obtained by both the PBPK approach and allometric scaling of a PPK model were compared a posteriori to the paediatric study observations. Both PPK and PBPK approaches allowed an adequate prediction of the PK of ivabradine and its metabolite in children. The second work was done after the study conduction in the paediatric population. The study was performed in 116 children (74 received ivabradine, 42 received the placebo) aged from 6 months to less than 18 years old and data were analysed. The relationship between blood and plasma concentrations was described using linear mixed effect models. In order to describe ivabradine and its metabolite blood concentrations in children, a joint population PK model was developed taking into account weight & age effects on PK parameters. Plasma exposure comparison indicated that higher dose/kg were necessary to achieve a similar exposure between younger and older children. The PK/PD relationship in adult patients is conserved in children.

Gliomas are the most common primary brain tumours and are associated with poor prognosis. Among them, diffuse gliomas – which include their most aggressive form glioblastoma (GBM) – are known to be highly infiltrative. The diagnosis and follow-up of gliomas rely on positron emission tomography (PET) and magnetic resonance imaging (MRI). However, these imaging techniques do not currently allow to assess the whole extent of such infiltrative tumours nor to anticipate their preferred invasion patterns, leading to sub-optimal treatment planning. Mathematical tumour growth modelling has been proposed to address this problem. Reaction-diffusion tumour growth models, which are probably the most commonly used for diffuse gliomas growth modelling, propose to capture the proliferation and migration of glioma cells by means of a partial differential equation. Although the potential of such models has been shown in many works for patient follow-up and therapy planning, only few limited clinical applications have seemed to emerge from these works. This thesis aims at revisiting reaction-diffusion tumour growth models using state-of-the-art medical imaging and data processing technologies, with the objective of integrating multi-parametric PET/MRI data to further personalise the model. Brain tissue segmentation on MR images is first addressed with the aim of defining a patient-specific domain to solve the model. A previously proposed method to derive a tumour cell diffusion tensor from the water diffusion tensor assessed by diffusion-tensor imaging (DTI) is then implemented to guide the anisotropic migration of tumour cells along white matter tracts. The use of dynamic [S-methyl-11C]methionine ([11C]MET) PET is also investigated to derive patient-specific proliferation potential maps for the model. These investigations lead to the development of a microscopic compartmental model for amino acid PET tracer transport in gliomas. Based on the compartmental model results, a novel methodology is proposed to extract parametric maps from dynamic [11C]MET PET data using principal component analysis (PCA). The problem of estimating the initial conditions of the model from MR images is then addressed by means of a translational MRI/histology study in a case of non-operated GBM. Numerical solving strategies based on the widely used finite difference and finite element methods are finally implemented and compared. All these developments are embedded within a common framework allowing to study glioma growth in silico and providing a solid basis for further research in this field. However, commonly accepted hypothesis relating the outlines of abnormalities visible on MRI to tumour cell density iso-contours have been invalidated by the translational study carried out, leaving opened the questions of the initialisation and the validation of the model. Furthermore, the analysis of the temporal evolution of real multi-treated glioma patients demonstrates the limitations of the formulated model. These latter statements highlight current obstacles to the clinical application of reaction-diffusion tumour growth models and pave the way to further improvements. / Les gliomes sont les tumeurs cérébrales primitives les plus communes et sont associés à un mauvais pronostic. Parmi ces derniers, les gliomes diffus – qui incluent la forme la plus agressive, le glioblastome (GBM) – sont connus pour être hautement infiltrants. Le diagnostic et le suivi des gliomes s'appuient sur la tomographie par émission de positons (TEP) ainsi que l'imagerie par résonance magnétique (IRM). Cependant, ces techniques d'imagerie ne permettent actuellement pas d'évaluer l'étendue totale de tumeurs aussi infiltrantes ni d'anticiper leurs schémas d'invasion préférentiels, conduisant à une planification sous-optimale du traitement. La modélisation mathématique de la croissance tumorale a été proposée pour répondre à ce problème. Les modèles de croissance tumorale de type réaction-diffusion, qui sont probablement les plus communément utilisés pour la modélisation de la croissance des gliomes diffus, proposent de capturer la prolifération et la migration des cellules tumorales au moyen d'une équation aux dérivées partielles. Bien que le potentiel de tels modèles ait été démontré dans de nombreux travaux pour le suivi des patients et la planification de thérapies, seules quelques applications cliniques restreintes semblent avoir émergé de ces derniers. Ce travail de thèse a pour but de revisiter les modèles de croissance tumorale de type réaction-diffusion en utilisant des technologies de pointe en imagerie médicale et traitement de données, avec pour objectif d'y intégrer des données TEP/IRM multi-paramétriques pour personnaliser davantage le modèle. Le problème de la segmentation des tissus cérébraux dans les images IRM est d'abord adressé, avec pour but de définir un domaine propre au patient pour la résolution du modèle. Une méthode proposée précédemment permettant de dériver un tenseur de diffusion tumoral à partir du tenseur de diffusion de l'eau évalué par imagerie DTI a ensuite été implémentée afin de guider la migration anisotrope des cellules tumorales le long des fibres de matière blanche. L'utilisation de l'imagerie TEP dynamique à la [S-méthyl-11C]méthionine ([11C]MET) est également investiguée pour la génération de cartes de potentiel prolifératif propre au patient afin de nourrir le modèle. Ces investigations ont mené au développement d'un modèle compartimental pour le transport des traceurs TEP dérivés des acides aminés dans les gliomes. Sur base des résultats du modèle compartimental, une nouvelle méthodologie est proposée utilisant l'analyse en composantes principales pour extraire des cartes paramétriques à partir de données TEP dynamiques à la [11C]MET. Le problème de l'estimation des conditions initiales du modèle à partir d'images IRM est ensuite adressé par le biais d'une étude translationelle combinant IRM et histologie menée sur un cas de GBM non-opéré. Différentes stratégies de résolution numérique basées sur les méthodes des différences et éléments finis sont finalement implémentées et comparées. Tous ces développements sont embarqués dans un framework commun permettant d'étudier in silico la croissance des gliomes et fournissant une base solide pour de futures recherches dans le domaine. Cependant, certaines hypothèses communément admises reliant les délimitations des anormalités visibles en IRM à des iso-contours de densité de cellules tumorales ont été invalidée par l'étude translationelle menée, laissant ouverte les questions de l'initialisation et de la validation du modèle. Par ailleurs, l'analyse de l'évolution temporelle de cas réels de gliomes multi-traités démontre les limitations du modèle. Ces dernières affirmations mettent en évidence les obstacles actuels à l'application clinique de tels modèles et ouvrent la voie à de nouvelles possibilités d'amélioration. / Doctorat en Sciences de l'ingénieur et technologie / info:eu-repo/semantics/nonPublished