The effect of nutrient, biodiesel and biochar amendments on the biodegradation of pollutant hydrocarbons was investigated in soil spiked with crude oil and coal tar over periods of up to 180 days. Biodiesel was chosen as an amendment for increasing the bioavailability of hydrocarbon pollutants because of its good solvent properties but low toxicity. Results from laboratory microcosm experiments on soil spiked with weathered and un-weathered tar showed that after 60 days, the concentrations of the potent carcinogen benzo[a]pyrene (BaP) were significantly reduced by 92 and 81%, respectively, in the biodiesel amended samples compared to the 17 and 26% reduction in the controls, and 8 and 34% depletion observed in the nutrient-only amended microcosms, respectively. The 3-ring PAH anthracene was also almost completely biodegraded in all the biodiesel amended experiments. However, phenanthrene degradation was significantly inhibited in these samples as only 0-2% reduction occurred after 180 days as opposed to the losses (>70%) observed in the control and the nutrient amended experiments. Apparent increases in concentrations were observed for some 4-ring PAHs, while the degradation of other 5 and 6-ring compounds was enhanced in the biodiesel amended samples probably due to their increased solubilisation by biodiesel. A stepwise treatment approach conducted on tar spiked soil revealed a higher reduction in BaP (98%) in the biodiesel amended microcosms compared to the control (29%) and phenanthrene depletion was also enhanced by 51% after 60 days of adding biodiesel to soil initially treated with nutrients. A similar trend in PAH degradation was observed for the crude oil spiked soil, but in these experiments the removal of the n-alkanes was significantly enhanced by nutrient amendment alone, while degradation of the branched alkanes was increased in the biodiesel treatments. Toxicity assays showed that biodiesel amended microcosms stimulated phosphatase enzyme activity and exhibited a lower toxic response to Microtox Vibrio fischeri. In this study, biochar amendment did not significantly reduce residual pollutant concentrations. Overall, the pattern observed in the removal of the PAHs using biodiesel, suggests the cometabolic action of ligninolytic fungi, probably via lignin peroxidases, as also evidenced from the visible growth of moulds after 7-14 days of amendment. The enhanced removal of carcinogenic PAH and the reduced toxicity observed in soil after biodiesel amendment, indicates that this bioremediation technique has potential for full scale field trials.

The expanded bed adsorption (EBA) process naturally assumes a mixed bed (MB) in terms of particle size and displays particle dispersion. To date models have not accounted for this simultaneously and accurately. Moreover, there has not been any work to predict an adsorption response when the bed is in a hydrodynamic-transient state which occurs for example while changing feedstock. A series of MB steady state hydrodynamic EBA models were developed which are increasingly close mimics of reality. This was achieved by progressively considering: a single representative equivalent particle size per axial position (MBEQD), separate particle size categories by using size-partition (MBSP) approach and inclusion of an additional component flux due to particle dispersion. Breakthrough predictions using the MB approach were more accurate compared to both that of a mono-sized bed and perfectly classified bed approaches clearly demonstrating its importance. The results of both MBEQD and MBSP were in close agreement with 40 cm bed height breakthrough experimental data. An important weakness in an existing method of including the particle dispersion was identified and a model for its more accurate representation developed. A transient hydrodynamic EBA model was developed by integrating the mono-sized transient hydrodynamic model and adsorption model. A simulation study using this demonstrated the possibility of loading while a bed is still expanding which may afford an increase in an operational throughput. The effects of various physical parameters on the performance of EBA were investigated using simulation. Windows of operation in relation to fluid velocity and load volume were determined which would satisfy minimum yield and throughput criteria. The model was also used to determine optimal loading time strategies in order to maximise yield and throughput. Finally a preliminary work explaining future potential developments in EBA modelling was performed.

Lentiviral vectors used in clinical trials are currently produced by transient transfection of adherent human embryonic kidney (HEK)293(T) cells. However, this approach is not scalable and for commercialisation the development of alternative strategies based on suspension-adapted producer cell lines, that have all genes for vector production stably integrated, is desired. To assist progress in this area, the aim of this thesis was to establish a microscale cell culture platform that enables key bioprocess design data to be acquired rapidly and cost-effectively. ProSavin®, an equine infectious anaemia virus (EIAV)-derived lentiviral vector developed for the treatment of Parkinson’s disease (Palfi et al., 2014) was used as a model system. First, the suitability of a shaken 24-well plate system for the suspension culture of HEK293T-derived producer cells was established. This system was shown to support equivalent cell growth and ProSavin® titres to conventional shake flasks while providing substantially greater opportunity for parallelisation. Second, the utility of the microscale platform when combined with statistical Design of Experiments (DoE) techniques for optimising titres and informing the design of a scale-up strategy was demonstrated. An initial screening experiment identified three parameters as having a critical influence on ProSavin® titres, which were post-induction period, liquid fill volume and concentration of doxycycline (inducer compound). Subsequent optimisation experiments defined operating ranges for these parameters. Third, the insights obtained during the microwell investigations were shown to aid successful scale-up of the ProSavin® process to a single-use 2 L WAVE bioreactor. Fourth, with a view to informing further improvements in process design, the half-life of ProSavin® and other EIAV-based lentiviral vectors was determined, and approaches to moderate the rate of decay during upstream processing trialled. Overall, it was concluded that the microwell platform may be viewed as an effective tool for future use in the development of lentiviral vector bioprocesses.

Atherosclerosis is a focal inflammatory disease, non-uniformly distributed in regions of disturbed flow within the arterial system. The local accumulation of lipids and inflammatory cells in these regions drives lesion initiation and progression into advanced plaques. Along with systemic risk factors, local haemodynamic metrics, such as wall shear stress (WSS), have been shown to be good correlates to plaque development. Few studies have investigated the relationship between mechanics and biology longitudinally over plaque development. However, conflicting views on the role of low and high wall shear stress in the development of the disease remain. Using high-resolution imaging and novel techniques like 3-D histology, we seek to provide more conclusive evidence of the correlation between perturbed haemodynamics and markers of atherosclerosis in two longitudinal studies in different animal models. The first study comprised of the development of a computational workflow to analyse co-localisations of selected disease markers with perturbed shear regions, induced in carotid arteries of ApoE-/- mice. Serial, in vivo mouse-specific haemodynamics were computed in micro-CT based geometries and coupled with plaque distribution spatially mapped on the in vivo lumen at each time point. Two novel haemodynamic metrics and other established metrics were assessed to find the best predictor of local areas of plaque formation. A quantitative analysis revealed plaque lipids and macrophages in the initial stage of the disease best overlapped with our custom metric, the low shear index (LSI), suggesting the importance of low shear in the initiation of atherosclerosis. In the second study, the developed workflow was applied to correlate plaque biology and mechanical metrics in transgenic hypercholesterolemic Yucatan minipigs. Lesions were induced by implanting a stenotic stent in one of the coronaries, and in vivo haemodynamics were computed in OCT based coronary reconstructions serially from healthy to advanced disease stages. Animals developed advanced human-like atherosclerotic lesions in regions of sustained low and oscillatory shear, induced by the stent. Histology based macrophage distribution showed good correlations with the newly developed metric, LSI, further emphasising the importance of low shear in atherogenesis. In conclusion, we have developed an integrated platform for high-resolution studies of plaque mechanobiology. Low shear was the best correlate to markers of atherosclerosis in two animal models, quantified by our novel metric-LSI, which might be a good predictor of disease development. Additionally, this thesis shows the importance of longitudinal in vivo studies of atherosclerosis and contributes to further understanding of this disease, while also providing tools that can be further extended in future studies.

Recent years have seen considerable interest in the use of microfabricated systems in chemical and biological due to their significant advantages in terms of speed, analytical throughput, yield, unit cost, footprint, reagent requirements and control. Inevitably this has led to a growing interest in transferring chromatographic methods to planar chip formats since techniques using high performance LC play such a prominent role in modern bioanalysis. In addition, the manipulation of multiphase (or segmented) flows within microfluidic channels has been recently investigated as a promising approach for large-scale experimentation in biology and chemistry. Importantly, flow segmentation allows for the compartmentalisation of reagent volumes ranging from a few femtolitres to hundreds of nanolitres within a continuous and immiscible fluid, the production of monodisperse droplets at high frequencies, the accurate control of droplet contents and the ability to perform kinetic analysis with high precision. Accordingly the integration of droplet-based microfluidics with HPLC has the potential to dramatically reduce dispersion and minimise dead volume effects by using droplets to collect fractions of the column effluent. This basic progress preserves the chemical identity of each fraction allowing further analysis downstream. In this work, microfluidic devices were fabricated using thermoset polyester (TPE) to operate under high pressure which is required for LC separation and high frequency droplet generation. The optical characteristics of the fabricated devices were assessed for feasibility of optical detections for droplets. Substrate resistance to pressure also was investigated for droplet generation with high frequency. Lastly, droplets were generated under various conditions by adjusting flow-rates and the oil viscosity. Secondly LC separation columns were formed in TPE channels using two different column materials: particulate and polymer monolithic columns. The packed channels were investigated by SEM. In addition, permeability was calculated from back25 pressures measured as a function of flow-rates and compared with columns. Neurotransmitters were separated by the columns to estimate performance. Thirdly, the both operations, LC separation and droplet-based microfluidics, were combined in a single planar format. Sequential operations of separation, compartmentalisation and concentration gradient generation were integrated on a single chip and characterised using confocal laser-induced fluorescence detection. Finally, a preliminary investigation is reported into the possibility of the indirect electrochemical detection as a universal detection that can monitor electrochemically detectable samples as well as non- or less-electroactive bio samples. Amino acids were separated by a commercial RPHPLC column and detected indirectly.

This thesis presents a musculoskeletal model that predicts the muscle and joint forces in the upper limb during an extreme activity. The scapula is an important link in the kinematic and dynamic chain of the upper limb; with its muscles acting as the primary stabilisers to the inherently unstable glenohumeral joint, thus allowing effective transmission of load through the kinematic chain of the shoulder. This bone is poorly represented in musculoskeletal models during these activities. Large soft-­‐tissue artefacts are a key reason for this. The shoulder is particularly prone to injury in overhead activities of the upper limb. Heavily loaded activities in these positions are of interest because they represent a limit, in that few people are capable of performing them. Pull-­‐ups are a common training activity that involve the movement of a large load with the arms overhead. Predicting the forces involved in such an activity allows a testing of current model limits and hypotheses on the function and biomechanics of the scapula. A novel methodology to track the dynamically moving scapula is validated using motion capture technology. This method is shown to improve measurement accuracy when compared to the literature. Kinematics of the scapula and upper limb are thus measured, presented and discussed for three types of pull-­‐up activity. The modelling aspects of the work build on a previous upper limb model, primarily adapting the kinematics representation. This better respects the measured kinematics through a relaxation of the closed-­‐chain mechanism as well as improving the ability to non-­‐homogeneously scale the model. The inverse dynamics description is modified to allow a variable hand load, muscle wrapping parameters and changed to prevent sudden unphysiological changes in moment arms and muscle bounds are increased to allow equilibrium to be reached with the inter-­‐segmental moments. Musculoskeletal loads are thus presented using a model that allows the dynamic analysis of extreme activities. Eccentric loading of the supraspinatus, deltoid and triceps was found to exist in potentially vulnerable positions, coinciding with a high incidence of impingement injury in pull-­‐up type activities. The glenohumeral joint reaction force is seen to be more centralised with a general increase in rotator cuff activation, although teres major and posterior deltoid seem to be key stabilisers. Pectoralis major was detrimental to stability, highlighting the importance of the scapula in positioning muscles during overhead activities. Comparison of model predictions with literature EMG results show good agreement.

Previous work has shown that nitric oxide production alters the relative height of the dicrotic notch (RHDN) in the peripheral arterial pulse of rabbits (Weinberg et al., 2001) and people (Chowienczyk et al., 1999), most likely through an influence on wave reflection. Preliminary evidence in both species shows that administration of glucose also alters notch height or augmentation index (AI), putatively because glucose stimulates the release of insulin which in turn stimulates the release of nitric oxide. This phenomenon could form the basis of a simple, non-invasive test of endothelial function. The work described in this thesis was designed to assess further the effect of glucose on arterial wavespeed (as a measure of stiffness), wave reflection and waveform. In the first study, an MR scanner was used to obtain brachial artery diameters and blood flow rates through the cardiac cycle in healthy volunteers before and after oral administration of glucose. In preliminary trials, MR images were processed in MATLAB® using different image analysis techniques; normalised cross correlation (NCC), a well-established method, was chosen for the automated acquisition of diameter and velocity waveforms. Pulse wave behaviour was then analysed using the diameter-velocity form of Wave Intensity Analysis (WIA) method, recently developed by Feng and Khir (2010). The study failed to show the expected effects of glucose. Furthermore, the diameter distentions detected for the brachial artery during the cardiac cycle appeared unphysiologically high and the calculated wave speeds appeared unphysiologically low. There were three possible sources of error: (1) the image analysis technique used to determine diameter waveforms, (2) the derivation of the new diameter-velocity form of WIA, and (3) the sugar hypothesis itself. The rest of the thesis describes investigations of these potential problems. To check the NCC-based image analysis methods, they were used in a study of aortic MR data that had previously been analysed by other methods (Li et al., 2010). Similar to the brachial artery/sugar experiment, non-physiological wavespeed values and high diameter distenstions were obtained, presumably due to an inaccuracy of the NCC image analysis technnique. Additionally, the wave behaviour in different aortic locations did not agree with well-established properties, presumably due to errors in the new WIA method. An attempt was made to find a consistent relation between NCC-derived diameters and manually-derived diameters, in order to develop a method for correcting the NCC data. However, no consistent relationship was found. This study provided additional evidence that the NCC method used in conjunction with noisy MRI data is unreliable. To assess the sugar-vasodilatation hypothesis, volume waveforms in the finger and pressure waveforms in the radial artery were measured by, respectively, photoplethysmography (PPG) and applanation tonometry, before and after the administration of a glucose or control drink to healthy volunteers. The radial pressure waveforms were used to calculate central aortic pressure waveforms, using the generalised transfer function described by Karamanoglu et al. (1993). Al, RHDN and the relative height of the diastolic peak (RHDP) of these waveforms were analysed. Despite the small sample size, a significant result was obtained for PPG RHDN, and borderline trends were obtained for other indices. Overall, the results were interpreted as supporting the hypothesis. Finally, the reliability of using the new diameter-velocity method of WIA was compared with the original pressure-velocity method in a polyurethane model of the aorta and its major branches. Pressure, velocity and diameter along the aorta to the femoral artery, and along arm vessels, were measured in the model and wave behaviour in the brachial artery and aorta was examined using both forms of WIA. Additionally, the aorta of the model was wrapped with tape or cling film in order to alter its compliance, an effect that was expected in the real circulation after the administration of glucose. Effects of aortic stiffening on wavespeed and wave intensity were observed in the aorta and arm arteries. However, the diameter-velocity form of WIA gave results that disagreed strongly with theoretical predictions. There may therefore be practically insupportable assumption in its derivation. In conclusion, it appears that both the NCC-MRI method and the diameter-velocity theory used in the initial brachial artery study may be unreliable. Some evidence was obtained for the hypothesised effect of sugar using other techniques.

The mouse brain is a highly specialised machine where each component plays a very specific role. We find organised structures and patterns everywhere, with columnar and layered design being a hallmark of brain cortex. Within this topographically organised neural networks we find that their components have well-defined roles. This specificity is reflected on their morphology, electrophysiology, connectivity pattern, and relative location. In order to fully understand these networks it is necessary to exploit all information available. Extracellular recordings are one of the most widely used techniques by neuroscientists, specially when interested in retrieving information from large populations of neurons with submillisecond precision. However, spatial information and cell classification are very limited and existing methods do not take advantage of the information available from high density electrode arrays recordings. In this thesis, I introduce two novel models that exploit spatial information conveyed by the extracellular signal recorded on this type of arrays. A simple model to localise the neural soma in three dimensional space and a model to parametrise salient morphological features of neurons. These models provide information that will prove useful when studying detailed organisations in neural networks. The localisation and morphological of neurons requires models that captures the features that impact the pattern induced at short recording distances as those achieved on high density electrode array recordings. Both localisation and classification must be tested for different morphological classes, as found in cortex. In order to evaluate the models we must first generate realistic simulated data on which models could be tested. Localisation must hold irrespective of neuronal type while classification is dependant on neuronal type. The different simulated neurons should reflect the different signal patterns seen on real recordings. After validate the models on realistic simulated data we have to evaluate the models on real recordings. The localisation algorithm successfully recovers the position of simulated neurons with low errors for distances within expected ranges in cortex. When localising neurons from real recordings we obtained a distribution of positions that agree with expected ranges in cortex. We separated simulated morphological classes using our classification model, validating the model as a tool to identify morphological classes from extracellular recordings. We also tested this model on real recorded data and using its parameters we identified putative morphological classes. We then verified these classes had different response properties to stimuli and firing patterns, supporting our theory that by using the amplitude pattern of the extracellular potential we can identify different neuronal types. In this thesis we combined the recovered spatial information with response properties from neighbouring neurons, to characterise the functional topographic organisation in deep layers of core auditory cortex. Finding a fractured representation, with local populations having similar response properties and high signal correlation on average, but large differences in response properties were possible. In agreement with recent imaging studies from upper layers of mouse auditory cortex that report smooth tonotopy on a large scale, but fractured tonotopy on a fine scale.

This Thesis applies complex spatial filters to the front end filtering to a computer vision framework for object recognition and scene categorization. This involves careful filter design in the Fourier domain based on discrete frame properties. Biological plausibility of the suggested filtering is compared against a common model found in the computer vision literature. The designed complex filter bank is equipped with focus-of-attention operators. Specifically, two possible keypoint detection methodologies are examined and compared with state of the art keypoint detection methods. This includes an investigation of scale-estimation methods. In addition, three image patch descriptor arrangements are proposed to sample the complex filter responses, and an initial evaluation of categorization performance is undertaken. Next, the spatial pooling arrangement of the best performing descriptor is further optimised and the performance of different complex filter bandwidths is examined in class separation tasks. A further study is conducted on the effects of a Winner-Take-All (WTA) approach to modifying filter responses before pooling. A thorough evaluation of descriptor performance is undertaken to reveal any advantages or disadvantages from a variety of perspectives. Next, the clustering behaviour of descriptors of various types is inspected in the descriptor feature space. A reverse look-up of visual words attempts to relate clustering behaviour to descriptor performance. Typical grouping approaches, such as spatial pyramids, are then compared with a novel method for coupling visual words in which a linear kernel SVM learns class separability. A final evaluation on this stage is presented and discussed, leading to conclusive arguments about the importance of careful approaches to word-pairing for good-quality categorization.

The models of the microcirculation of blood and interstitial fluid in the human liver lobule are developed based on the classical hexagon model of Kiernan. Both blood and interstitial flows in the lobule are treated as flows in porous medium connected via the fenestrated membrane of sinusoids. Several important physiological components are developed and included in the models. The lobule with tissue elasticity shows that the pressure-flux relationship is non-linear and the poroelastic model has more compliance than the solid elastic model. Models of the interstitial flow in both a single lobule and the whole liver are also developed. The results show that our models can predict the amount of interstitial fluid drainage including the ascites. From the parameter studies, we find that the permeabilities of the sinusoids and the interstitial space, and the portal pressure are the most important factors on ascites production. We further investigate the oxygen transportation and uptake by liver cells using the advection–diffusion equations and Michaelis-Menten kinetics. The studies show that the main mechanism of oxygen transportation within the sinusoids is advection; however, the transportations within the interstitial space and across the fenestrated endothelial cells are mainly from diffusion process. The effect of the arrangement of the vessels and the geometry of the lobule on blood perfusion and oxygen distribution is also studied. The results show that the classical hexagonal lobule with the vascular septa provides the optimal perfusion compared to other geometries of the lobule. In summary, this thesis contributes to the development of mathematical models of several important features in the liver microcirculation such as the tissue elasticity, the interstitial flow, the oxygen distribution, and the arrangement of the vessels in the lobule.