Os métodos de controle de qualidade tradicionais aplicados ao radiodiagnóstico, é a melhor maneira de garantir a boa qualidade das imagens produzidas. No entanto, a investigação de particularidades oriundas do processo de formação de imagens radiológicas requer ferramentas computacionais complementares, em função do número de variáveis envolvidas. Entretanto, os fantomas computacionais baseados em voxels não conseguem representar as variações morfométricas necessárias para a simulação de exames cujo diagnóstico é baseado em imagem. Neste trabalho foi desenvolvido um novo tipo de fantoma computacional, baseado em modelagem 3D, que possui as vantagens apresentadas pelos fantomas computacionais tradicionais sem os problemas encontados nestes. A ferramenta de modelagem utilizada, o Blender, é disponibilizada gratuitamente na internet. A técnica utilizada foi a box modeling, que consiste na deformação de uma primitiva básica, nesse caso um cubo, até que apresente a forma da estrutura que se deseja modelar. Para tanto, foram utilizadas como referencia, imagens obtidas de atlas de anatomia e fotografias de um esqueleto fornecido pela Universidade de Mogi das Cruzes. Foram modelados o sistema ósseo, os órgãos internos e a anatomia externa do corpo humano. A metodologia empregada permitiu a alteração de parâmetros do modelo dentro da ferramenta da modelagem. Essa possibilidade foi mostrada através da variação, dos formatos do intestino e do aumento da quantidade de tecido adiposo da malha referente a pele. A simulação das imagens radiológicas foi realizada a partir de coeficientes de atenuação de massa de materiais, ossos e tecidos e de modelos com diversas características físicas. Essa versatilidade permite prever a influência que as diferenças morfométricas entre os indivíduos provocam nas imagens, propriciando dessa forma, uma ferramenta relevante complementar aos métodos de controle de qualidade tradicionais. / The conventional methods of quality control applied to radio diagnosis are the best way to have assured good quality of the produced images. Due the amount of variables to consider, the study of particular issues of the process of formation of radiological images requires complementary computational tools. However, the computational voxel based phantoms are not suitable to represent the morphometrical variations, intended for test simulations with image based diagnosis. This work developed a new type of computational phantom, based on 3D modelling. It has the same advantages of the conventional ones, without some of their restrictions. The modeling tool employed, Blender, is available on internet for free download. The project uses the technique called box modeling, which consists in the deformation of a primitive form (a cube, in this case) until it presents a similar form to that it is wanted to model. In order to achieve it, some images, obtained from anatomy atlas and a skeleton pictures obtained from University of Mogi das Cruzes, were used as reference. Were built models from skeletal system, internal organs and external human body anatomy. The applied methodology allowed model´s parameter settings on the modelling tool. This option was presented by means of intestine format variation and increase of adipose tissue on the mesh that represents skin. The simulation of radiological images was done by means of x-ray mass attenuation coefficients, bones and tissues and models with diferent physical characteristics. This flexibility allows the analysis and forecasting of the influences that morphometrical differences of individual implies on images, revealing an important tool that complements the conventional quality control tools.

Musical dynamics- loudness and changes in loudness - forms one of the key aspects of expressive music performance. Surprisingly this rather important research area has received little attention. A reason is the fact that while the concept of dynamics is related to signal amplitude, which is a low-level feature, the process of deriving perceived loudness from the signal is far from straightforward. This thesis advances the state of the art in the analysis of perceived loudness by modelling dynamic variations in expressive music performance and by studying the relation between dynamics in piano recordings and markings in the score. In particular, we show that dynamic changes: a) depend on the evolution of the performance and the local context of the piece; b) correspond to important score markings and music structures; and, c) can reflect wide divergences in performers' expressive strategies within and across pieces. In a preparatory stage, dynamic changes are obtained by linking existing music audio and score databases. All studies in this thesis are based on loudness levels extracted from 2000 recordings of 44 Mazurkas by Frederic Chopin. We propose a new method for efficiently aligning and annotating the data in score beat time representation, based on dynamic time warping applied to chroma features. Using the score-aligned recordings, we examine the relationship between loudness values and dynamic level categories. The research can be broadly categorised into two parts. The first investigates how dynamic markings map to performed loudness levels. Empirical results show that different dynamic markings do not correspond to fixed loudness thresholds. Rather, the important factors are the relative loudness of neighbouring markings, the inter-relations of nearby markings and other score information, the structural location of the markings, and the creative license exercised by the performer in inserting further interpretive dynamic shaping. The second part seeks to determine how changes in loudness levels map to score features using statistical change-point techniques. The results show that significant dynamic score markings do indeed correspond to change points. Furthermore, evidence suggests that change points in score positions without dynamic markings highlight structurally salient events or events based on temporal changes. In a separate bidirectional study, we investigate the relationship between dynamic mark- ings in the score and performed loudness using machine learning techniques. The techniques are applied to the prediction of loudness levels corresponding to dynamic markings, and to the classification of dynamic markings given loudness values. The results show that loudness values and markings can be predicted relatively well when trained across recordings of the same piece, but fail dismally when trained across a pianist's recordings of other pieces. The findings demonstrate that score features may trump individual style when modelling loudness choices. The analysis of the results reveal that form|such as the return of the theme - and structure - such as repetitions -influence predictability of loudness and markings. This research is a first step towards automatic audio-to-score transcription of dynamic markings. This insight will serve as a tool for expression synthesis and musicological studies.

The process of a single cell evolving into a complex organism results from a series of coordinated movements of cells and tissues, especially during early embryo development. Although a wealth of morphological data characterises the shapes and movements of cells in embryos, how these movements are driven, patterned and controlled, and how this patterning is related to the mechanical properties of tissues remains unknown. Four-pole electromagnetic tweezers have been developed to probe the mechanical properties of living embryonic tissues that are undergoing active morphogenetic development. The device is capable of generating magnetic forces in the order of nano-Newtons on a grafted magnetic bead. The local passive mechanical properties of the tissues can be characterised by measuring the three-dimensional bead movement and analysing cell shape changes and cell rearrangement in response to this externally applied force. The magnetic device is used to probe the rheology in early zebrafish embryos between high stage (3.3 hpf) and the onset of gastrulation (5.3 hpf) when rapid cell cycles give way to a hollow sphere of cells. The tissue response to the applied force is modelled as linear visco- elastic. The embryo becomes stiffer and more viscous during this period of development, showing that a loose collection of cells becomes cohesive tissues. A computational model is used to explore how cells respond to local or global mechanical perturbations in two systems. First, the model simulates the movement of the bead within an embryo, and the results illustrate the generation, patterning and relaxation of the local cell stress around the bead. Second, the model reproduces the autonomous changes in mitotic cells within a stretched monolayer, and the results show that propensity of cells to divide along their long axis facilitates stress relaxation and contributes to tissue homoeostasis.

Sellafield is home to the UK's largest repository of nuclear waste, including reprocessed uranium and plutonium, as well as a backlog of unprocessed used fuel and waste kept in outdated storage facilities; commonly referred to as "legacy waste". For this reason, Sellafield has often been called the most hazardous place in Western Europe and as such, is currently undergoing a multi-billion pound decommissioning and clean-up operation. Each on-site facility has unique challenges associated with it, many of them presenting situations where the radiation chemistry aspects of the material degradation are not well understood. The key factors that can affect water radiolysis processes in the Sellafield challenges are a high pH environment, the presence of magnesium hydroxide, the presence of iron oxide, and the presence of organic materials. This work examines the effect each of these factors has on H2 and H2O2 production in water radiolysis as well as developing a computational model to offer some understanding to the kinetic behaviour of water radiolysis under such conditions. The computational model was able to replicate experimental measurements of radiolytic H2 and H2O2 production in both aerated and deaerated water at neutral pH, and provide a further understanding of the role of dissolved oxygen in water radiolysis. Measurements of H2O2 from solutions containing NaOH have shown that an increase in pH generally results in a higher steady state of H2O2, while measurements of H2 show a similar increase with a maximum production rate at pH ~11. The model was also able to closely replicate these experimental measurements with some over prediction, which highlights a gap in our understanding of high pH radiolysis and also brings into question the validity of the estimated rate constant for the reaction: O- + O2- → 2OH- + O2 k= 6.0×10^8 M^-1 s^-1 which was originally determined from kinetic model calculations designed to describe the decay of ozonide (O3ˉ) during pulse-radiolysis studies of high pH solutions conducted byK. Sehested et al in 1982.The radiolysis of magnesium hydroxide slurry also resulted in an increased yield of hydrogen gas but had little effect on the yield of hydrogen peroxide. The hydrogen yield was 0.52 molecules per 100eV while a NaOH solution of equivalent pH gave a yield of 0.27, however interference from carbonate may be the cause of the increased yield. A surface effect was also estimated to contribute 0.05 molecules per 100 eV to the hydrogen gas yield. Hydrogen gas and hydrogen peroxide was measured from the radiolysis of aqueous methanol. This was modelled with a near agreement, but modifications to the model were necessary; highlighting areas of the model that need improvement, as well as providing a reaction scheme from which a more comprehensive model for aqueous methanol radiolysis could be developed.

Antibodies and antibody fragments are the largest class of biotherapeutics in development with many products already available in the clinic. Antibodies are promising due to their naturally high affinity and specificity for biological targets. A key stumbling block to biopharmaceutical development compared to small molecule drugs is the general requirement for a stable liquid formulation, which is often difficult to obtain due to issues with aggregation, phase separation, particle formation, and chemical instabilities. Aberrant solution behaviour limits the production, storage and delivery of the monoclonal antibody. Biopharmaceutical solution behaviour is determined by weak, transient protein-protein and protein-solvent interactions. An attractive interaction potential between proteins in solution can lead to association. Irreversible association occurs when proteins undergo large scale structural changes and aggregate. Reversible association is less severe, but can lead to undesirable solution properties such as high viscosity, phase separation and opalescence, which can lead to difficulties throughout the downstream processing and formulation steps. These problems can become exacerbated during formulation of antibodies when trying to achieve high protein concentrations often required for effective antibody dosage. Firstly, we studied the domains of the Fab fragment using statistical models and continuum electrostatic calculations and found that the CH1 domain is more soluble than the other domains and has properties of intrinsically disordered like proteins which is supported by observations in the literature. We then investigated the immunoglobulin superfamily and found 11 proteins which may have a similarly disordered nature. We present a new web server for predicting protein solubility from primary sequence using an in-house algorithm that weighs the contribution of various sequence properties for predicting solubility. Lastly, we conducted physical characterisation of an antibody and human serum albumin in pharmaceutically relevant buffers and found that the interaction potential can be modelled using spherical models from low to high protein concentration. We hope that the work outlined in this thesis will contribute to the theoretical understanding and modelling of protein solution behaviour.

Lipases are enzymes that play fundamental roles in fat digestion and metabolism, and function at the interface formed between hydrophobic molecules and the surrounding aqueous environment. These interfacial interactions are thought to induce conformational changes in a "lid" region of the lipase, leading to a dramatic increase in activity. This thesis aims to provide insight into the interactions that govern lipase association with interfaces of di erent structural characteristics, and the possible conformational changes that arise as a function of these interactions. A multi-scale molecular simulation approach (combining atomistic and coarse-grained methods) was applied to study two different lipases with a range of interfaces, including "soft" biological surfaces and "hard" non-biological surfaces. Three major insights were gained from these studies. First, interactions of a small bacterial lipase (M37) with lipid interfaces resulted in substantial structural changes in a lid region, uncovering of the underlying active site. A mechanism of interfacial ac- tivation is proposed for this lipase. Second, the interaction of M37 with non-biological interfaces di er from lipid interfaces, leading to altered interfacial orientations with possible functional consequences. Third, the amino acid composition of the lid region of a yeast lipase (TLL) is shown to play crucial roles in lipase activation and structural stability.

Depression is a highly prevalent clinical condition which has been estimated to affect a growing part of the population in western countries. Alongside expenditure on diagnostics and treatment, there is a high economic impact due to lost productivity. Although a range of treatments are available, diagnoses are currently costly and require subjective assessment by a specialist. Moreover, treatment selection can be lengthy and can involve trial and error. To develop better diagnostics, stratification, and treatments for depression, we need a better understanding of the condition across different levels - from neural mechanisms to cognition and behaviour. Computational modelling is an emergent theory-driven approach which can aid linking data across different levels of analysis - from neural mechanisms and computations in the brain, to cognitive algorithms and observable behaviour. Some models integrate diverse findings and make predictions, while others enable inference of clinical measures which are not obvious in raw data. Modelling can lead to better understanding of depression, and in turn to better stratification and treatments. On the other hand, machine learning and classification methods can help detect clinically-relevant patterns in experimental data in a purely data-driven manner. This can lead to development of better screening and diagnostic methods. In the current work, we first review some of the most prominent neurocognitive theories of depression, as well as existing studies which used computational modelling methods. Based on our review, we argue that modelling can provide a rich set of tools for a better understanding of the condition. We then develop two novel computational modelling accounts of depression. In the first account, we propose an explicit mechanistic link between a robust behavioural negative bias effect and some of the widely reported or theorised neural aspects of depression - hyperactive amygdala and inhibited dopamine release. In the second account, we attempt to better explain depressive cognitive deficits and show how they can arise from depression-relevant etiological factors - altered valuation and controllability estimates. Finally, in the third part of this work we attempt to develop a novel system for detecting depressive symptoms based on a combination of face-tracking, eye-tracking and cognitive performance measures. We evaluate the system in a pilot experiment and show that a combination of measures can achieve better results than measures from each domain separately.

Thesis (Ph.D. (Physics)) -- University of Limpopo, 2018 / Beyond conventional intercalation chemistry, reaction of lithium with sulphur and oxygen (so-called “Li-air” batteries) have the potential to provide 2 to 5 times the energy density of current Li-ion battery systems. However, both Li/S and Li/O2 systems suffer from cycling performance issues that impede their commercial applications: Li/O2 cycling is limited by electrolyte decomposition and large cell polarization; Li/S suffers from the low conductivity of S and the solubility of intermediary polysulfide species during cycling. It has been reported that Se and mixed SexSy represent an attractive new class of cathode materials with promising electrochemical performance in reactions with both Li and Na ions. Notably, unlike existing Li/S batteries that only operate at high temperature, these new Se and Li/SexSy electrodes are capable of room temperature cycling. Initially, stabilities of insoluble discharge products of oxygen and sulphur in the Li-S and Li-O2 batteries were investigated using density functional theory within the generalized gradient approximation, and these were deduced from their structural, electronic and mechanical properties. The structural properties are well reproduced and agree to within 3% with the available experimental data. Li2S, Li2O and Li2O2 and Li2S2 structures all have negative heats of formations indicating that they are stable, however, that of Li2S2 structure was relatively high compared to others. Calculated phonon dispersion and elastic properties revealed that Li2O, Li2S and Li2O2 structures are mechanically stable and great agreement with experimental work. The Li2S2 structure displayed soft modes associated mainly with sulphur atoms vibrations in the a-b plane, hence it is not mechanically stable in agreement with the negative C13. Stable Li2S2 polymorphs were extracted from soft modes of calculated phonon dispersions along the gamma direction in the Brillioun zone. Temperature is known to have a significant impact on the performance, safety, and cycle lifetime of lithium-ion batteries (LiB). In order to explore properties of discharge products associated with Li/S and Li/Se batteries at different temperatures, molecular dynamics and cluster expansion methods were employed. The former was achieved by firstly deriving empirical interatomic potentials of Li2S and Li2Se which were fitted to experimental and DFT calculated data. The potentials were validated against available experimental and calculated structure, elastic properties and phonon spectra. In addition, complex high temperature transformations and melting of Li2S and Li2Se were reproduced, as deduced from molecular dynamics simulations. Both Li2S and Li2Se were found to withstand high temperatures, up to 1250K each which is a desirable in future advanced battery technologies. Furthermore, cluster expansion and Monte-Carlo simulations were employed to determine phase changes and high temperature properties of mixed Li2S-Se. The former generated 42 new stable multi-component Li2S-Se structures and ranked metastable structures by enthalpy of formation. Monte Carlo simulations produced thermodynamic properties of Li2S-Se system for the entire range of Se concentrations obtained from cluster expansion and it demonstrated that Li2S-Se is a phase separating system at 0K but changes to mixed system at approximately 350K which was confirmed by constructed by phase diagram of Li2S-Se system. It was finally demonstrated that molecular dynamics and Monte Carlo simulations techniques yield consistent results on phase separation and high temperature behavior of Li2S-Se at 50% of sulphur and selenium.

Optical tweezers using highly focussed laser beams can be used to exert forces and torques and thus drive micromachines. This opens up a new field of microengineering, whose potential has yet to be fully realized. Until now, methods that have been used for modelling optical tweezers are limited to scatterers that are homogeneous or that have simple geometry. To aid in designing more general micromachines, I developed and implemented two main methods for modelling the micromachines that we use. These methods can be used for further proposed structures to be fabricated. The first is a FDFD/T-matrix hybrid method that incorporates the finite difference frequency domain (FDFD) method, which is used for inhomogeneous and anisotropic media, with vector spherical wave functions (VSWF) to formulate the T-matrix. The T-matrix is then used to calculate the torque of the trapped vaterite sphere, which is apparently composed of birefringent unit crystals but the bulk structure appears to be arranged in a sheaf-of-wheat fashion. The second method is formulating the T-matrix via discrete dipole approximation (DDA) of complex arbitrarily shaped mesoscale objects and implementing symmetry optimizations to allow calculations to be performed on high-end desktop PCs that are otherwise impractical due to memory requirements and calculation time. This method was applied to modelling microrotors. The T-matrix represents the scattering properties of an object for a given wavelength. Once it is calculated, subsequent calculations with different illumination conditions can be performed rapidly. This thesis also deals with studies of other light scattering phenomena including the modelling of scattered fields from protein molecules subsequently used to model FRET resonance, determining the limits of trappability, interferometric Brownian motion and the comparison between integral transforms by direct numerical integration and overdetermined point-matching.

Judgements of and sensitivity to style are ubiquitous. People become sensitive to the structural regularities of complex or “polymorphous” categories through exposure to individual examples, which allows them respond to new items that are of the same style as those previously experienced. This thesis investigates whether a dimension reduction mechanism could account for how people learn about the structure of complex categories. That is, whether through experience, people extract the primary dimensions of variation in a category and use these to analyse and categorise subsequent instances. We used Singular Value Decomposition (SVD) as the method of dimension reduction, which yields the main dimensions of variation of pixel-based stimuli (eigenvectors). We then tested whether a simple autoassociative network could learn to distinguish paintings by Picasso and Braque which were reconstructed from only these primary dimensions of variation. The network could correctly classify the stimuli, and its performance was optimal with reconstructions based on just the first few eigenvectors. Then we reconstructed the paintings using either just the first 10 (early reconstructions) or all 1,894 eigenvectors (full reconstructions), and asked human participants to categorise the images. We found that people could categorise the images with either the early or full reconstructions. Therefore, people could learn to distinguish category membership based on the reduced set of dimensions obtained from SVD. This suggests that a dimension reduction mechanism analogous to SVD may be operating when people learn about the structure and regularities in complex categories.