The Vestibulo-Ocular Reflex (VOR) produces compensatory eye movements in response to head and body rotations movements, over a wide range of frequencies and in a variety of dimensions. The individual components of the VOR are separated into parallel pathways, each dealing with rotations or movements in individual planes or axes. The Horizontal VOR (hVOR) compensates for eye movements in the Horizontal plane, and comprises a linear and non-linear pathway. The linear pathway of the hVOR provides fast and accurate compensation for rotations, the response being produced through 3-neuron arc, producing a direct translation of detected head velocity to compensatory eye velocity. However, single neurons involved in the middle stage of this 3-neuron arc cannot account for the wide frequency over which the reflex compensates, and the response is produced through the population response of the Medial Vestibular Nucleus (MVN) neurons involved. Population Heterogeneity likely plays a role in the production of high fidelity population response, especially for high frequency rotations. Here we present evidence that, in populations of bio-physical compartmental models of the MVN neurons involved, Heterogeneity across the population, in the form of diverse spontaneous firing rates, improves the response fidelity of the population over Homogeneous populations. Further, we show that the specific intrinsic membrane properties that give rise to this Heterogeneity may be the diversity of certain slow voltage activated Potassium conductances of the neurons. We show that Heterogeneous populations perform significantly better than Homogeneous populations, for a wide range of input amplitudes and frequencies, producing a much higher fidelity response. We propose that variance of Potassium conductances provides a plausible biological means by which Heterogeneity arises, and that the Heterogeneity plays an important functional role in MVN neuron population responses. We discuss our findings in relation to the specific mechanism of Desynchronisation through which the benfits of Heterogeneity may arise, and place those findings in the context of previous work on Heterogeneity both in general neural processing, and the VOR in particular. Interesting findings regarding the emergence of phase leads are also discussed, as well as suggestions for future work, looking further at Heterogeneity of MVN neuron populations.

Cette thèse traite de la modélisation des réacteurs chimiques par la « modélisation compartimentale », qui consiste à diviser le système en un réseau d’une dizaine à quelques centaines de volumes interconnectés, appelés compartiments. La structure du réseau est déduite à partir d’informations provenant d’expériences de traçage, d’informations techniques sur le réacteur chimique, de simulations de mécanique des fluides numérique et des objectifs de la modélisation. Cette méthode procure un bon compromis entre temps de calcul et finesse des résultats. Quand ils sont correctement menés, les modèles à compartiments donnent des prédictions similaires, en termes de réactions chimiques, à ceux issus des simulations de mécanique des fluides numérique réactive avec un temps de calcul plus court et une représentation physique plus concrète du comportement du réacteur. Chaque étude issue de la littérature est consacrée à un réacteur spécifique avec une approche particulière qui ne peut pas être directement transposée sur un autre réacteur. L’objectif de cette thèse est d’apporter une contribution au développement d’une méthodologie la plus générale possible et de développer un outil de génération automatique et de résolution du système d’équations différentielles qui doit être résolu. Dans le premier chapitre, un état de l’art est réalisé, définissant le champ d’application de notre méthode, dans le but d’identifier les méthodes de découpage les plus pertinentes et les différentes méthodes pour calculer les échanges entre les compartiments. Dans un second chapitre, une méthode générale pour de la modélisation compartimentale est développée. Une approche polyvalente est proposée, consistant à découper le réacteur en tranches identiques. Le calcul des échanges entre compartiments, dus à la convection et la turbulence, est présenté en détail, avec la description des trois méthodes de calcul des échanges turbulents. Une interface a été développée permettant de construire n’importe quel réseau de compartiments. À partir de cette interface, les équations sont écrites et automatiquement résolues. La méthode est appliquée dans un troisième chapitre sur un cas défavorable au découpage en tranches. Cela a permis de tester les limites de cette approche. En particulier, deux points ont été étudiés : (1) l’applicabilité du découpage en tranches identiques et (2) la comparaison entre les méthodes de calcul des échanges turbulents. Le premier test a prouvé la robustesse de l’approche par division mais le second test n’a pas permis d’établir si une méthode de calcul est meilleure qu’une autre. Finalement, la méthode a été valorisée et transférée en implémentant les algorithmes développés dans un logiciel commercial. Ce logiciel permet de simuler la dispersion d’espèces réactives et non réactives (traceurs), dans un modèle contenant plusieurs centaines de compartiments organisés en tranches identiques / This PhD deals with modelling of chemical reactors with the “compartmental modelling” approach, which consists in dividing the system into a network from a dozen to several hundreds of interconnected volumes, called compartments. The structure of the network is deduced from tracer experiments, technical information about the chemical reactor and computational fluid dynamics flow simulations. This method provides a good compromise between computation time and results accuracy. When they are properly set-up, compartmental models give similar predictions, in terms of chemical reactions, as those of CFD simulations with a shorter calculation time and a more concrete representation of the reactor behavior. Every study from the literature is devoted to a specific reactor with a particular approach that cannot be straightforwardly transposed to other reactors. The aim of this PhD is to provide a contribution to the development of the most general possible methodology and to develop an automatic tool of generation and resolution of the differential equations system which must be solved. In the first chapter, a state of the art is proposed, defining the field of application of our method, in order to identify the most relevant division methods and the different methods to calculate the exchange between compartments. In the second chapter, a general methodology for compartmental modelling is developed. A versatile approach is proposed, consisting in dividing the reactor in identical slices. The calculation of exchange between compartments, both due to convection and turbulence, is presented in detail, with the description of three calculation methods for turbulent exchange. An interface has been developed, allowing to build any network of compartments. From this interface, the equations are written and solved automatically. The methodology is applied in the third chapter to an unfavorable case for slice cutting. This has allowed to test the limit of this approach. In particular, two points have been studied: (1) the applicability of division into identical slices and (2) the comparison between the turbulent exchange calculation methods. The first test has proved the robustness of the division approach but the second test has not allowed to establish whether one calculation method is better than another. Finally, the methodology has been promoted and transferred by implementing the developed algorithms within a commercial software. This software allows to simulate the dispersion of reactive and non-reactive (tracers) species, in model containing hundreds of compartments organized in identical slices

In translational medicine, complementary functional and morphological imaging techniques are used extensively to observe physiological processes in vivo and to assess structural changes as a result of disease progression. The combination of magnetic resonance imaging (MRI) and positron emission tomography (PET) provides excellent soft tissue contrast from MRI with exceptional sensitivity and specificity from PET. This thesis explores the use of sequentially acquired PET and MR images to improve the quantification of small animal PET data. The primary focus was to improve image-based estimates of the arterial input function (AIF), which defines the amount of PET tracer within blood plasma over time. The AIF is required to produce physiological parameters quantifying key processes such as metabolism or perfusion from dynamic PET images. The gold standard for AIF measurement, however, requires serial blood sampling over the course of a PET scan, which is invasive in rat studies but prohibitive in mice due to small total blood volumes. To address this issue, the geometric transfer matrix (GTM) and recovery coefficient (RC) techniques were applied using anatomical MR images to enable the extraction of partial volume corrected image based AIFs from mouse PET images. A non-invasive AIF extraction method was also developed for rats, beginning with the optimization of an automated voxel selection algorithm to assist in extracting MR contrast agent signal time courses from dynamic susceptibility contrast (DSC) MRI data. This procedure was then combined with dynamic contrast enhanced (DCE) MRI to track a combined injection of Gadolinium-based contrast agent and PET tracer through the rat brain. By comparison with gold standard tracer blood sample data, it was found that normalized MRI-based AIFs could be successfully converted into PET tracer AIFs in the first pass phase when fitted with gamma variate functions. Finally, a MR image segmentation method used to provide PET attenuation correction in mice was validated using the Cambridge split magnet PET/MR scanner?s transmission scanning capabilities. This work recommends that contributions from MR hardware in the PET field of view must be accounted forto gain accurate estimates of tracer uptake and standard uptake values (SUVs). This thesis concludes that small animal MR data taken in the same imaging session can provide non-invasive methods to improve PET image quantification, giving added value to combined PET/MR studies over those conducted using PET alone.