Search

This project aimed to investigate the forces that both needle and tissue experienced during a needle insertion, and how they altered the needles trajectory. An investigation into the current literature showed that existing skin tissue surrogates did not perform similarly to real skin tissue in vivo during needle insertions. A new surrogate is required to aid with validation for computational models of needle insertions, while avoiding the ethical issues raised from testing real tissue. This study developed an improved skin tissue surrogate for use in photoelastic testing which focused on replicating the fracture mechanism observed during a needle insertion through human skin tissue. It is demonstrated that konjac glucomannan gel fractures in the same way as human skin tissue. Experimental assessments determined that at a concentration of 1.5% gel powder to water konjac jelly had a stiffness which closely matched the stiffness of human skin tissue in vivo. In order to use the surrogate in photoelastic analysis it must be clear and exhibit temporary birefringence, and it is shown that with careful preparation konjac satisfies these criteria. The strain optic coefficient for the gel is determined, which links the optical response to the strain and stress experienced by the surrogate. A variety of needle insertion experiments were conducted which assess how varying the insertion speed, needle length, and needle gauge affect the overall response. The results prove that konjac jelly accurately replicates needle insertion response through soft tissue better than existing surrogates. With use of the GFP2500 poleidoscope, a novel digital polariscope, full field and directional information from a needle insertion is obtained. The results identify never-before-seen locations of principal strain magnitude near the puncture surface. For the first time the forces directional response was reported, and show how a bending moment acts on the needle; resulting in deflection.

Accurate and reliable registration of shapes and multi-dimensional point sets describing the morphology/physiology of anatomical structures is a pre-requisite for constructing statistical shape models (SSMs) and atlases. Such statistical descriptions of variability across populations (regarding shape or other morphological/physiological quantities) are based on homologous correspondences across the multiple samples that comprise the training data. The notion of exact correspondence can be ambiguous when these data contain noise and outliers, missing data, or significant and abnormal variations due to pathology. But, these phenomena are common in medical image-derived data, due, for example, to inconsistencies in image quality and acquisition protocols, presence of motion artefacts, differences in pre-processing steps, and inherent variability across patient populations and demographics. This thesis therefore focuses on formulating a unified probabilistic framework for the registration of shapes and so-called \textit{generalised point sets}, which is robust to the anomalies and variations described. Statistical analysis of shapes across large cohorts demands automatic generation of training sets (image segmentations delineating the structure of interest), as manual and semi-supervised approaches can be prohibitively time consuming. However, automated segmentation and landmarking of images often result in shapes with high levels of outliers and missing data. Consequently, a robust method for registration and correspondence estimation is required. A probabilistic group-wise registration framework for point-based representations of shapes, based on Student’s t-mixture model (TMM) and a multi-resolution extension to the same (mrTMM), are formulated to this end. The frameworks exploit the inherent robustness of Student’s t-distributions to outliers, which is lacking in existing Gaussian mixture model (GMM)-based approaches. The registration accuracy of the proposed approaches was quantitatively evaluated and shown to outperform the state-of-the-art, using synthetic and clinical data. A corresponding improvement in the quality of SSMs generated subsequently was also shown, particularly for data sets containing high levels of noise. In general, the proposed approach requires fewer user specified parameters than existing methods, whilst affording much improved robustness to outliers. Registration of generalised point sets, which combine disparate features such as spatial positions, directional/axial data, and scalar-valued quantities, was studied next. A hybrid mixture model (HMM), combining different types of probability distributions, was formulated to facilitate the joint registration and clustering of multi-dimensional point sets of this nature. Two variants of the HMM were developed for modelling: (1) axial data; and (2) directional data. The former, based on a combination of Student’s t, Watson and Gaussian distributions, was used to register hybrid point sets comprising magnetic resonance diffusion tensor image (DTI)-derived quantities, such as voxel spatial positions (defining a region/structure of interest), associated fibre orientations, and scalar measures reflecting tissue anisotropy. The latter meanwhile, formulated using a combination of Student’s t and Von-Mises-Fisher distributions, is used for the registration of shapes represented as hybrid point sets comprising spatial positions and associated surface normal vectors. The Watson-variant of the HMM facilitates statistical analysis and group-wise comparisons of DTI data across patient populations, presented as an exemplar application of the proposed approach. The Fisher-variant of the HMM on the other hand, was used to register hybrid representations of shapes, providing substantial improvements over point-based registration approaches in terms of anatomical validity in the estimated correspondences.

Cardiovascular diseases are a significant health risk worldwide, being the largest contributor to deaths in most developed and developing countries. For physical testing of new medical devices, diseased tissue specimens are desirable. However, these are difficult to obtain in quantity. The first phase of the work therefore focused on emulating effects of disease on artery mechanical response, using enzyme and chemical treatment of healthy tissue. Porcine aorta was partially digested by elastase and collagenase treatments to remove constituent proteins, and exposed to low concentration glutaraldehdye to partially cross-link proteins. Uniaxial tension testing and controlled peel testing were then performed in the artery axial and circumferential directions to assess the changes in mechanical and failure behaviour. The treatments successfully altered the wall tensile and peeling response with effects varying by the loading type and direction. Multiphoton microscopy was also performed to allow visualisation of the changes to fibre structure and density. Finally, tensile test results were fitted to the Gasser-Ogden-Holzapfel constitutive model and a continuum damage model, and the fitted curves were best matched with the circumferential direction results. The latter phase of the work focused on development of methods for characterising and simulating catheter-induced dissection processes. An experimental procedure was developed, wherein a catheter was forced between layers of arterial media, propagating a dissection, while reaction force was measured. The various approaches utilised to model this process within FEA are presented and the subsequent difficulties explored. The inherent complexity of the process being modelled resulted in difficulty drawing out the underlying problems. To rectify this, the experiment was simplified such that a metal wedge with a rounded front was used to dissect the tissue. This was successfully modelled and insights from this were considered with regard to numerical difficulties in the catheter dissection model.

Personalised medicine is redefining the present and future of healthcare by increasing treatment efficacy and predicting diseases before they actually manifest. This innovative approach takes into consideration patient’s unique genes, environment, and lifestyle. An essential component is physics-based simulations, which allows the outcome of a treatment or a disease to be replicated and visualised using a computer. The main requirement to perform this type of simulation is to build patient-specific models. These models require the extraction of realistic object geometries from images, as well as the detection of diseases or deformities to improve the estimation of the material properties of the studied object. The aim of this thesis was the design of a general framework for creating patient- specific models for biomechanical simulations using a framework based on statistical shape models. The proposed methodology was tested on the construction of spine models, including vertebrae and intervertebral discs (IVD). The proposed framework is divided into three well-defined components: The paramount and first step is the extraction of the organ or anatomical structure from medical images. In the case of the spine, IVDs and vertebrae were extracted from Magnetic Resonance images (MRI) and Computed Tomography (CT), respectively. The second step is the classification of objects according to different factors, for instance, bones by its type and grade of fracture or IVDs by its degree of degeneration. This process is essential to properly model material properties, which depends on the possible pathologies of the tissue. The last component of the framework is the creation of the patient-specific model itself by combining the information from previous steps. The behaviour of the developed algorithms was tested using different datasets of spine images from both computed tomography (CT) and Magnetic resonance (MR) images from different institutions, type of population and image resolution.