The measurement of ankle joint dorsiflexion is an important component of the clinical examination of the foot and is also an outcome measure often employed in research. Diagnosis of ankle equinus or limited ankle dorsiflexion is based solely on this measurement. Although a great majority of research papers in this field utilise normal clinical goniometers for obtaining the maximum ankle dorsiflexion angle (MADA) and important clinical decisions are based on these findings, there is overwhelming evidence that this method is highly unreliable. Thus this thesis aimed to investigate alternative methods of measuring ankle dorsiflexion and various issues that affect the MADA in order to obtain a clear picture of foot segment movement from which an Ankle Goniometer could be consequently designed and validated. Ten different techniques were identified that included various apparatuses designed specifically for measuring ankle dorsiflexion. However, during validation, their methodological quality would have benefitted from the use of an actual patient population and comparison with a reference standard, which caused papers to score poorly on methodological quality assessment. It was concluded that issues affecting this measurement and that needed to be researched further include: foot posture, the amount of moment applied and the stretching characteristics of the calf muscle tendon unit. Experiments within this study indicated that in adults, the mean maximum pronated angles were always higher than in other postures, with 8.27o difference between the pronated and supinated postures (p=0.032) and 5.78o between pronated and neutral (p=0.000). However, the reported difference between neutral and supinated (mean angle of 2.49o) is of little clinical significance. It was observed that the forefoot always travels through a greater angle than the hindfoot in all 3 foot postures, while the hindfoot to forefoot angle increases during the application of a moment, indicating that the ‘midtarsal joint locking mechanism’ cannot be applied to passive dorsiflexion of the foot. In adolescents, there are no significant differences in the mean MADA between the 3 postures, implying that this measurement technique may be performed in any of these postures. It was also concluded that there is no need to control moment applied during this measurement procedure to produce consistent results between raters and that the calf muscle tendon unit does not stretch significantly following brief repetitive passive stretching. ii An ankle goniometer that measured purely hindfoot movement by eliminating forefoot influence and that increased reliability by holding the foot in the chosen posture, was consequently designed. This was validated by synchronizing with an electrogoniometer and an optoelectronic motion capture system. Reliability testing, with the foot held in a supinated posture as opposed to the traditional Rootian method of placing the foot at subtalar joint neutral position, spanned a number of different trials, including intra-tester and inter-tester reliability studies utilising both controlled and uncontrolled conditions, in convenience samples of healthy participants and a random sample of patients. Reliability testing between 4 raters with little experience on the utilization of this device was finally conducted. It has been shown that the Ankle Goniometer is a valid and reliable device for measuring ankle joint complex dorsiflexion both in healthy adults and in a patient population.

For several years, microwave engineers have dreamed of using non-ionising electromagnetic waves in medical imaging applications. The rate of evolution of microwave techniques for medical application has been immense and shows no signs of relenting. Since the limitations of X-ray mammography are well reported, alternative techniques using microwaves for breast cancer detection are developed. Some of these techniques have progressed to the point where positive clinical experience has emerged.

The natural response to the intrusive bodily sensation of discomfort is positional change. The purpose of this study was to explore how children and young people with profound physical, learning and communication difficulties, largely dependent on others to gauge their need for positional change, have their comfort needs met when using postural management equipment. Thirteen qualitative case studies were undertaken. Nine of the participants attended a special needs education primary or secondary school, two were in transition to school and two attended day services. All participants had a neurodevelopmental disability, with each being the focus of one case study. Parents, teachers, therapists and key support staff were interviewed, and the school or day centre routines of the children and young people were observed, with selective video recording. Single case and cross case analyses were undertaken. The findings showed threats to comfort include the restrictive nature of various accessories, hastiness of care tasks producing positioning errors and the procedural stretching of tissues prior to application. Opportunities for lessening discomfort included scheduled daily routines and time out of postural management equipment. Equipment use for the children and young people in this study was intrinsically coupled with care giving. Attentive caregivers read the behavioural expressions of the children and young people and reassuringly responded, safeguarding them from discomforting experiences. This maximised each individual‘s adaptive functioning, without compromising the benefits of postural support. This study highlights that the same item of equipment can be both comfortable and uncomfortable. Given the social and interactional world in which the children and young people live and learn, and the complex nature of their difficulties, it is others who must accept responsibility for ensuring their optimal level of comfort.

This thesis presents research and development in registration of images containing rigid structures using deformable models. In the existing methods of deformable image registration, the feature based methods using purely interpolation are often used under the assumption that all parts of the image can be deformed non-linearly. However, it is not applicable for some images, such as medical images which contain some bony structures. Therefore, it is more reasonable in image registration to treat the parts representing rigid structures differently from the remaining parts. A novel method is proposed in this thesis that uses a flexible spring-mass system to provide a good representation of image information. Anatomical structures are represented by spring masses and their interconnecting springs located along the structure boundaries. Hence, the system can treat some pre-selected parts of an image as rigid structures, whilst it is possible to elastically deform the rest of the image. The deformation caused by the displacement of the features is modelled by the spring-mass system via the motion equation. To ensure high quality registration, the image content is introduced as external forces acting on the masses and is iteratively employed into the motion equation to achieve a better and better registration result (It should be noted that the improvement in registration quality is limited by the quality and content of original images). The performance of the proposed method is shown using simulated data, and is validated using real Computed Tomography (CT) images. Objective measures and images that enable subjective evaluation of the new technique are provided. The proposed method is fast, robust and offers high registration accuracy. This is due to the following developments: • Innovatory approach based on the spring-mass system for image registration. • A new spring force formula to support the new method. • External forces based on image content are employed and applied via an iterative approach to ensure the high quality registration. • A fast initialisation method for the spring-mass system generation. • A fast interpolation technique to reduce the computation time in the iterative registration stage.

Automated classification of medical images for colorectal and prostate cancer diagnosis is a crucial tool for improving routine diagnosis decisions. Therefore, in the last few decades, there has been an increasing interest in refining and adapting machine learning algorithms to classify microscopic images of tumour biopsies. Recently, multispectral imagery has received a significant interest from the research community due to the fast-growing development of high-performance computers. This thesis investigates novel algorithms for automatic classification of colorectal and prostate cancer using multispectral imagery in order to propose a system outperforming the state-of-the-art techniques in the field. To achieve this objective, several feature extraction methods based on image texture have been investigated, analysed and evaluated. A novel texture feature for multispectral images is also constructed as an adaptation of the local binary pattern texture feature to multispectral images by expanding the pixels neighbourhood to the spectral dimension. It has the advantage of capturing the multispectral information with a limited feature vector size. This feature has demonstrated improved classification results when compared against traditional texture features. In order to further enhance the systems performance, advanced classification schemes such as bag-of-features - to better capture local information - and stacked generalisation - to select the most discriminative texture features - are explored and evaluated. Finally, the recent years have seen an accelerated and exponential rise of deep learning, boosted by the advances in hardware, and more specifically graphics processing units. Such models have demonstrated excellent results for supervised learning in multiple applications. This observation has motivated the employment in this thesis of deep neural network architectures, namely convolutional neural networks. Experiments were also carried out to evaluate and compare the performance obtained with the features extracted using convolutional neural networks with random initialisation against features extracted with pre-trained models on ImageNet dataset. The analysis of the classication accuracy achieved with deep learning models reveals that the latter outperforms the previously proposed texture extraction methods. In this thesis, the algorithms are assessed using two separate multiclass datasets: the first one consists of prostate tumour multispectral images, and the second contains multispectral images of colorectal tumours. The colorectal dataset was acquired on a wide domain of the light spectrum ranging from the visible to the infrared wavelengths. This dataset was used to demonstrate the improved results produced using infrared light as well as visible light.

Due to geometrical features, it is claimed that larger femoral heads in total hip replacement (THR) are superior in achieving normal biomechanics than smaller ones; and that hip resurfacing (RSF) is superior to THR. This has not been conclusively proven. Most studies have investigated level walking, which may not be demanding enough to highlight what could be small biomechanical differences between implants. Few biomechanical studies have compared more demanding tasks and not with patients with different femoral head sizes or RSF. This thesis aimed to address these omissions by investigating level walking, stair descent and sit-to-stand (STS) biomechanics between three groups (32mm THR, 36mm THR and RSF). Twenty-six osteoarthritis patients were recruited and tested pre-operatively, then three and twelve months post-operatively. Demographic differences between groups were expected due to patient considerations for different implants, so a study was performed to determine whether level walking biomechanics alter progressively during the aging process with a group of 63 healthy participants. Three matched sub-groups were extracted from this group as controls. There was no suggestion that gait deteriorates progressively with age. Hip reconstruction, irrespective of head size, can allow patients to return to the biomechanical levels of controls during level walking. Stair descent differences remained 12 months post-operatively in cadence (p=0.042) and peak hip power generated (p<0.001) compared to controls. The 32mm group exhibited vertical ground reaction force (vGRF) asymmetry pre-operatively (p<0.001) and 3 months post-operatively (p=0.013); and impulse asymmetry (p<0.001) pre-operatively during STS. The 36mm group exhibited impulse asymmetry (p=0.05)three months post-operatively. This thesis is the first biomechanical analysis of stair descent and STS of two THR groups and a RSF group. It has demonstrated stair descent differences at 12 months post-operatively and overloading of the healthy limb in some THR patients. The latter could be problematic for the healthy limb.

Controlled by extracellular signals, tumour-induced angiogenesis is a crucial step in the development of tumours. Among the many cell signals already identified, the VEGF and Notch signalling pathways play a critical role in controlling endothelial cells (ECs) during angiogenesis. Although this regulatory mechanism has become a current research focus in biology, its computational modelling is still rare. We focus on developing a computational model to simulate the VEGF and Notch signalling regulatory mechanism to perceive the micro procedure of angiogenesis in silico and fill the gap between biology and computer engineering. We first developed a mathematical model with nonlinear partial differential equations (PDEs) to describe the migration of endothelial tip cells during tumour-induced angiogenesis. The simulation results show that both chemotaxis and haptotaxis have impacts on the migration of ECs in velocity and density, and the impacts depend on the gradient and direction of tumour angiogenenic factor (TAF), and fibronectin, implying a possible malignant mechanism for some subgroups of tumour. We then developed the model further to simulate the regression, recurrence or clearance of tumours due to tumour cytotoxic factors, including the immune system and drugs delivered through the vessels formed during angiogenesis, providing a broader understanding of tumours. Based on the PDE model which provided parameters of continuum mathematical model, we finally developed an enzymatic catalysed regulating model in the form of ordinary differential equations (ODEs) with agent-based modelling (ABM) using Java and MATLAB languages, to visually realise the sprouting regulated by VEGF and Notch signalling during angiogenesis. The simulation describes the process of how an endothelial stalk cell becomes an endothelial tip cell, and sprouts under the influence of VEGF and Notch signalling, revealing the relationship between sprouting and branching. As the simulation results are consistent with reported in vitro and in vivo assays, the study bridges angiogenesis research and computer modelling from the dynamic regulatory mechanism perspective, offering a huge leap over previous studies in computationally simulating tumour-induced angiogenesis. It is hoped that the results will assist researchers in both the experimental and theoretical angiogenesis communities to improve understanding of the complexity and identify the fundamental principles of angiogenesis, whilst also using modelling approaches that will enrich knowledge for computational scientists in this field.

In total knee replacement (TKR) surgery, mobile bearing (MB) total knee prostheses were designed to more closely mimic the function of the normal knee than traditional fixed bearing (FB) designs by allowing axial mobility between the polyethylene insert and tibial tray. Despite the hypothetical benefits of the MB design, few studies have objectively analysed knee biomechanics during activities of daily living (ADLs) in the laboratory compared to FB designs. This thesis aimed to substantiate the theoretical advantages of MB implantation during ADLs in the laboratory as well as during free living conditions, in addition to investigating previous claims of instability in MB knees. Sixteen patients undergoing primary unilateral total knee replacement (TKR) surgery were randomised to receive either a FB (n=8) or MB (n=8) total knee prosthesis and were tested at pre-surgery, three months post-surgery, and nine months post-surgery using three dimensional motion analysis in the laboratory and electrogoniometry and accelerometry during free living conditions. No differences were found between FB and MB groups during walking at post-surgery that could not be explained by differences at pre-surgery. There were also no differences between FB and MB groups during the more biomechanically demanding activities of stair negotiation and sit to stand and stand to sit activities, as well as no differences during free living conditions away from the laboratory. There appears to be no evidence based rationale for the widespread use of MBs with regards to optimising knee function during ADLs. This thesis was the first to compare FB and MB designs using the same implant range, posterior cruciate ligament (PCL) scenario, posterior stabilising strategy, and patella strategy over a range of ADLs, as well as being the first to combine testing in the laboratory with testing during free living conditions away from the laboratory.

The aim of this thesis is to develop a theoretical model to explore how sociomaterial networks, involving large-scale automation, come into being, persist and change over time, within a healthcare environment. It does so by bridging the gap between design, implementation and use of large-scale pathology automation (LSPA) within two United Kingdom (UK) National Health Service (NHS) laboratories. A longitudinal, multi-site, ethnographic approach was used, along with semi-structured interviews, template analysis and participant observation of LSPA ‘in-practice’. This research has suggested that design features, embedded within the material properties of LSPA, were purposefully intended to bring about organisational change. In both user organisations, the material affordances of LSPA resulted in anticipated skill mix changes. However, material constraints required the enforcement of changes to organisational routines, creating operational difficulties, which were then subsequently transferred across organisational boundaries by the researcher/manager. The identification of these sociomaterial affordances and constraints, in conjunction with humans acting as boundary objects, had the unintended consequence of influencing strategic decision making and initiating structural and cultural change. The development and practical application of the resulting SociomANTerial model allowed the researcher to trace the analytical history of these organisational changes over time and consider the impact of broader social structures such as power. Ultimately it is suggested that a greater emphasis on collaboration between users, designers and corporate agents will result in more innovative approaches for technology adoption and improved organisational design.