Techniques described to date for the reproducible and noninvasive quantification of regional cardiac function have been largely relegated to research settings due to time-consuming and cumbersome image acquisition and analysis. In this thesis, feature tracking algorithms are developed for 2-D+Time cardiac magnetic resonance (CMR) and 3-D+Time cardiac computed tomography (CCT) image sequences that are easily acquired clinically, while emphasising reproducibility and automation in their design. First, a commercially-implemented CMR feature tracking algorithm for the analysis of steady state free precession (SSFP) cine series is evaluated in patients with hypertrophic cardiomyopathy (HCM) and arrhythmogenic right ventricular cardiomyopathy (ARVC), which primarily affect the left ventricle (LV) and right ventricle (RV), respectively, and functional impairment compared with control populations is found in both cases. The limitations of this implementation are then used to guide development of an automated algorithm for the same purpose, making use of fully convolutional neural networks (CNN) for segmentation and spline registration across all frames simultaneously for tracking. This study is performed in the subjects with HCM, and functional impairment is again identified in disease subjects. Finally, as myocardial contraction is inherently a 3-D phenomenon, a technique is developed for quantification of regional function from 3-D+Time functional CCT studies using simultaneous registration of automatically generated Loop subdivision surface models for tracking. This study is performed in canine mongrels, and compared with the current state of the art technique for CCT functional analysis. This work demonstrates the feasibility of automated, reproducible cardiac functional analysis from CMR and CCT image sequences. While work remains to be done in extending the principles demonstrated and modular components described to fully automated whole-heart analysis, it is hoped that this thesis will accelerate the clinical adoption of regional functional analysis.

This study develops a 3D meso-scale fracture characterisation and modelling framework for better understanding of the failure mechanisms in concrete, by combining in-situ micro-scale X-ray computed tomography (XCT) experiments and XCT image-based finite element (FE) simulations. Firstly, sophisticated in-situ XCT experiments are conducted on concrete cubes under Brazilian-like, uniaxial and cyclic compression. Proper procedures for XCT image reconstruction and multi-phasic segmentation are identified. The fracture evolution at different loading stages is characterised and visualised as well as the detailed distributions of aggregates and voids. The Young's moduli of aggregate and cement are obtained by micro-indentation tests and used in XCT-image based asymptotic homogenisation simulations to calculate effective elastic constants of concrete cubes. The XCT technique proves very powerful in characterising and visualising the complicated 3D fracture evolution in concrete. The material properties and the segmented 3D images from the experiments are then used for FE fracture simulations with realistic aggregates, cement and voids. Image-based mesh generation algorithms are developed for 2D in a MATLAB code and identified for 3D in Simpleware. Cohesive interface elements are embedded within cement and aggregate-cement interfaces to simulate the complex nonlinear fracture. Extensive simulations of 40mm and 20mm cubes under compression and tension are carried out. Good agreements are achieved between the load-displacement curves and final crack patterns from the simulations and those from the compressive in-situ XCT tests. The XCT image-based modelling proves very promising in elucidating fundamental mechanisms of complicated crack initiation and propagation in concrete.

Introdução: As reabsorções radiculares internas (RRI) são lesões de difícil prognóstico e o diagnóstico é baseado em exames clínicos e de imagem. Estudos que avaliam a acurácia dos exames por imagem para a detecção de RRI, utilizam simulações realizadas por brocas, criando cavidades com forma e limites definidos diferentes da realidade clínica. Além disso, sabe-se que tão importante quanto a detecção da lesão de RRI é a capacidade de visualização do envolvimento da estrutura dentinária adjacente. Assim, este trabalho objetivou desenvolver uma nova metodologia para simulação de RRI com base na desmineralização ácida e, após, testar a influência de diferentes protocolos de aquisição de imagens tomográficas na medição do volume dessas lesões. Materiais e métodos: Foi testada uma metodologia de simulação de RRI, com ação progressiva de ácido nítrico 5% e hipoclorito de sódio 8%, em onze dentes monorradiculares. As cavidades geradas foram avaliadas em Microscopia Eletrônica de Varredura (MEV) e as medidas de diâmetro e profundidade foram correlacionadas com os tempos de ação do ácido. Imagens de Tomografia Computadorizada de Feixe Cônico (TCFC) foram adquiridas de todos os dentes da amostra em dois aparelhos, de FOV amplo e restrito, com três protocolos de tamanho de voxel cada: I-Cat Next Generation (voxel 0,200; 0,250; 0,300 mm) e Kodak 9000 3D (voxel 0,076; 0,100; 0,200 mm). As simulações de RRI foram mensuradas com uma ferramenta de cálculo de volume do software Dolphin 3D. Os valores médios de volume das cavidades foram comparados com Análise de Variância (ANOVA) e Teste Tukey ( =0,05). A validação das mensurações de volume obtidas pelo software foi realizada por meio de moldagem das cavidades com silicona de adição. Resultados: As simulações de RRI reproduziram lesões de diferentes tamanhos e com profundidades de 0.22 mm a 1.59 mm. Em relação às medidas volumétricas, não houve diferença estatística significante (P<0,05) entre os volumes de RRI mensurados a partir dos dois menores tamanhos de voxel em cada aparelho de TCFC, mas ambos foram diferentes estatisticamente (P>0,05) do volume calculado com os maiores tamanhos de voxel. Ainda, as médias de volume a partir das impressões em silicona foram menores que as mensuradas pelo voxel 0,200 mm. Conclusão: A metodologia de simulação de RRI proposta gerou cavidades com limites irregulares e com razão de diâmetro:profundidade mais semelhante ao processo de RRI in vivo. Diferentes protocolos de TCFC influenciam nas medidas de volume de RRI simuladas, ressaltando a importância da padronização na aquisição das imagens para acompanhamento dessas lesões. / Introduction: The internal root resorption (IRR) injuries are difficult to predict and the diagnosis is based on clinical and radiographic examination. Studies that evaluated the accuracy of imaging tests for the detection of IRR using simulations produced by burs, created cavities with different forms and limits of clinical reality. Furthermore, it is known that as important as the IRR detection is the ability to preview the involvement of adjacent dentinal structure. Thus, this study aimed to develop a new methodology for IRR simulation-based with acid demineralization and after investigate the influence of different tomographic protocols in the measurement of the IRR lesions volume. Materials and methods: A methodology was desenvolved for simulation of IRR with progressive action of 5% nitric acid and 8% sodium hypochlorite in eleven teeth single-rooted. IRR cavities produced were evaluated by scanning electron microscopy (SEM), measured in diameter and depth and after correlated with the acid times. Cone Beam Computed Tomography (CBCT) images were acquired of all teeth in two tomography devices, with large and restrict field of view (FOV) using three voxel sizes protocols: I-Cat Next Generation (voxel 0.200, 0.250, 0.300 mm) and Kodak 9000 3D (voxel 0.076, 0.100, and 0.200 mm). IRR simulations were measured with a tool for calculating volume of the Dolphin 3D software. Volume means of the IRR cavities were compared using Analysis of Variance (ANOVA) and Tukey test ( = 0.05). A validation of the tomographic measurements was performed with addition silicone impression. Results: The simulations reproduced IRR lesions with different sizes and depths (0.22mm to 1.59mm). Regarding volumetric measurements, there was no statistically significant (P<0,05) difference between the IRR volumes measured from the two smaller voxel sizes in each CBCT device, but both were statistically different (P>0,05) from the calculated volume with larger voxel sizes. Still, the volume means from silicone impressions were lower than those measured at 0.200 mm voxel. Conclusion: The IRR simulation methodology generated cavities with irregular borders and diameter:depth ratio more similar to in vivo process of IRR. Different protocols of CBCT influences the volume measures of IRR simulated, emphasizing the importance of standardization in the acquisition of CBCT images for monitoring these lesions.

Value of HRCT in the diagnostics of pulmonary complications of malignant haematologic diseases. Early diagnosis is one of the essential conditions for successful therapy of heamatologic malignancies and their complications. Imaging methods play an important role in their detection and patient's management. HRCT is the most contributing method in the diagnostics of pulmonary involvement, which is very common in this group of patients. The aim of this thesis was to find out the potential of this method in the diagnostics of pulmonary complications in patients with haematologic malignancy. For this purpose were assessed 4 objectives: 1) comparison between HRCT and chest X-ray; 2) determination of capability of HRCT to make a specific diagnose; 3) definition of indication criteria for HRCT and 4) assignment of the suitable examination technique. In the time period from January 2000 until December 2005 were performed 214 HRCT examinations of the lungs in the group of 162 patients with haematological malignancy and pulmonary symptomatology. 176 examinations were compared with chest X-ray, which was performed 48 hours or less before HRCT. Pulmonary lesion was discovered in 180 cases (84,1%). 123 findings were confirmed by another method (response to a specific therapy, cultivation, serology, BAL or...

In 1996, the United States Food and Drug Administration officially approved the use of intravenous recombinant tissue-type plasminogen activator for treatment of acute ischemic stroke, with the requirement that a baseline computed tomography (CT) scan be performed to rule out acute intracerebral or subarachnoid hemorrhage. Today, the American Heart Association (AHA) Stroke Council acknowledges magnetic resonance imaging (MRI) as more sensitive to the detection of ischemia, and yet, guidelines released by the group suggest that either CT or MRI may serve as the primary, hyperacute imaging modality. The AHA recommends that for most cases, non-contrast-enhanced CT scans provide sufficient information for medical management decisions. A systematic review of published literature was conducted to compare current capabilities of CT and MRI in an effort to determine which imaging modality should be used in the setting of acute ischemic stroke. Current research indicates that MRI is comparable to CT in the detection of acute hemorrhage, but superior in the detection of acute ischemia. In addition, MRI has demonstrated the ability to not only identify suitable patients for treatment, but also identify patients whose treatment would be unnecessary and potentially dangerous. Therefore, the hope is that clinical guidelines, like those released by the AHA Stroke Council, will be modified to promote MRI as the primary imaging modality. A modification to the major clinical guidelines will initiate a change in the approach of acute stroke evaluation across all clinical stroke centers.

Osteoporosis is an age-related skeletal disease characterised by an increased incidence of fragility fractures. In this thesis I develop a new technique capable of measuring the thickness of the previously unmeasured endocortical region, and providing an improved measure of the cortical bone mineral density (cBMD) from in-vivo clinical CT scans. These features are of interest as both have been linked to fracture risk. \\ The new technique is developed within the cortical bone mapping (CBM) framework so that it provides localised architectural measurements over a bone's surface. Its performance is assessed using simulated QCT data from three simulated phantoms with differing bone architecture, and two paired datasets of ex-vivo QCT and HR-pQCT scans across the proximal femur and the lumbar spine. The simulated data allows for inaccuracies in CBM measurements caused by beam hardening effects to be considered for the first time: I show that beam hardening leads to an underestimation in cortical thickness and an overestimation in trabecular BMD and that these inaccuracies can be reduced through adjustments to the CBM optimisation process. \\ A new technique of analysing HR-pQCT scans is also developed, for the validation of the new CBM method. It was used in place of other established HR-pQCT techniques for its ability to provide localised measurements of the endocortical region. A comparison with the well known full-width half-maximum (FWHM) method shows that it is less susceptible to errors caused by beam hardening. It also measures the mean cBMD, which has a greater clinical relevance than the peak cBMD measured by the FWHM method as it includes the impact of porosity. I demonstrate that the endocortical thickness can be measured to an accuracy of \(-0.15\pm0.71\thinspace\mathrm{mm}\), and that local cBMD measurements are possible down to \(300\thinspace\mathrm{mg/cm}^3\) from QCT scans over the proximal femur. I also validate CBM methods over the vertebrae for the first time and demonstrate that the cortical thickness and endocortical thickness can be measured with accuracies of \(0.10\pm0.30\thinspace\mathrm{mm}\) and \(-0.20\pm0.53\thinspace\mathrm{mm}\). \\ Two clinical trials involving Teriparatide are used to demonstrate that the new CBM method is able to detect significant regional changes in the dense cortical and endocortical bone over the proximal femur and lumbar spine, which can be attributed to changes in intracortical remodelling and endosteal apposition. The analysis of a cross-sectional fracture discrimination trial shows that fracture incidence is associated with significant decreases in endocortical thickness over specific regions.

BACKGROUND: Cardiovascular disease, principally manifest as myocardial infarction or stroke, is the dominant cause of death worldwide and despite therapeutic advances, the global burden of these conditions continues to increase. In order to address this ongoing disease burden, there is a clear need to more effectively target the use of existing and novel diagnostic investigations and medical therapies. Emerging cardiovascular biomarkers include the biochemical, such as high-sensitivity cardiac troponin, and the radiological, such as computed tomography coronary angiography (CTCA) and 18Ffluoride positron emission tomography (PET). Cardiac troponins can now be reliably quantified in clinically stable or asymptomatic populations and provide information about myocardial pathophysiology, whilst CTCA can non-invasively quantify atherosclerotic burden and 18F-fluoride PET imaging offers insight into plaque vulnerability. Improved targeting of diagnostic investigations requires more reliable estimation of pre-test probability of coronary disease whilst optimizing the use of pharmacological or interventional treatments requires more accurate prognostic stratification. Achieving both objectives in an equitable manner across all population groups will depend upon updated clinical guidelines containing improved risk models and enhanced management pathways. The objective of this thesis was to investigate the potential clinical benefit of novel approaches to the diagnostic and prognostic assessment of coronary heart disease. EVALUATION OF THE 2016 NATIONAL INSTITUTE FOR HEALTH AND CARE EXCELLENCE (NICE) GUIDANCE ON THE ASSESSMENT OF SUSPECTED STABLE ANGINA. A post-hoc analysis was undertaken of the Scottish COmputed Tomography of the HEART (SCOT-HEART) trial of 4,146 participants with suspected angina randomised to assessment with computed tomography coronary angiography or standard care. Patients were dichotomised according to guideline definitions into groups representing possible angina and non-anginal presentations. The primary (diagnostic) endpoint was diagnostic certainty of angina at 6 weeks and the prognostic endpoint comprised fatal and non-fatal myocardial infarction. In 3,770 eligible participants, CTCA increased diagnostic certainty more in those with possible angina (relative risk [RR] 2.22 (95% CI 1.91-2.60), p < 0.001) than those with non-anginal symptoms (RR 1.30 (1.11-1.53), p=0.002; pinteraction < 0.001). In the possible angina cohort, CTCA did not change rates of invasive angiography (p=0.481) but markedly reduced rates of normal coronary angiography (hazard ratio [HR] 0.32 (0.19-0.52), p < 0.001). In the non-anginal cohort, rates of invasive angiography increased (HR 1.82 (1.13-2.92), p=0.014) without reducing rates of normal coronary angiography (HR 0.78 (0.30-2.05), p=0.622). At 3.2 years of follow-up, fatal or nonfatal MI was reduced in patients with possible angina (3.2% to 1.9%; HR 0.58 (0.34- 0.99), p=0.045) but not in those with non-anginal symptoms (HR 0.65 (0.25-1.69), p=0.379). Overall the updated NICE guidance on patient assessment maximises the benefits of CTCA with respect to diagnostic certainty, the use of invasive coronary angiography, and reductions in fatal and non-fatal myocardial infarction. Patients with non-anginal chest pain derive minimal benefit from CTCA, which instead increases rates of invasive investigation. EXTERNAL VALIDATION OF THE PROSPECTIVE MULTICENTER IMAGING STUDY FOR EVALUATION OF CHEST PAIN (PROMISE) TOOL FOR DETERMINING MINIMAL-RISK OF CORONARY ARTERY DISEASE. The PROspective Multicenter Imaging Study for Evaluation of chest pain (PROMISE) minimal-risk tool was recently developed to identify patients with suspected stable angina at very low risk of coronary artery disease and clinical events. The external validity of this tool was investigated within the context of the Scottish Computed Tomography of the HEART multicenter randomised controlled trial of patients with suspected stable angina due to coronary artery disease. Model discrimination and calibration was determined amongst 1,764 patients in whom complete CCTA data were available and compared with the European Society of Cardiology guideline-endorsed Coronary Artery Disease Consortium (CADC) risk score. The PROMISE minimal-risk tool improved discrimination compared with the CADC model (c-statistic 0.785 vs 0.730, p < 0.001) and was improved further following re-estimation of covariate coefficients (c-statistic 0.805, p < 0.001). Model calibration was initially poor (c2 197.6, Hosmer-Lemeshow [HL] p < 0.001), with significant overestimation of probability of minimal risk, but improved significantly following revision of the PROMISE minimal-risk intercept and covariate coefficients (c2 5.6, HL p=0.692). HIGH-SENSITIVITY CARDIAC TROPONIN I IN THE DIAGNOSIS OF STABLE CORONARY ARTERY DISEASE In a pre-specified sub-study of the Scottish COmputed Tomography of the Heart trial, plasma cardiac troponin was measured using a high-sensitivity single molecule counting assay in 943 adults with suspected stable angina who had undergone coronary computed tomography angiography. Rates of obstructive coronary artery disease were compared with the pre-test probability determined by the European Society of Cardiology Coronary Artery Disease Consortium risk model with and without cardiac troponin concentrations. External validation was undertaken in an independent study population from Denmark comprising 487 patients with suspected stable angina. Higher cardiac troponin concentrations were associated with obstructive coronary artery disease with a 5-fold increase across quintiles (9 to 48%, p < 0.001) independent of known cardiovascular risk factors (odds ratio [OR] 1.35 [95% confidence interval (CI) 1.25-1.46] per doubling of troponin). Cardiac troponin concentrations improved the discrimination of the ESC model for identifying obstructive coronary artery disease (c-statistic 0.785 to 0.800, p=0.003) and improved classification into ESCrecommended categories of clinical risk (net reclassification improvement 0.143 [95% CI, 0.093-0.193]). The revised model achieved similar improvements in discrimination and net reclassification when applied in the external validation cohort. HIGH-SENSITIVITY CARDIAC TROPONIN I IN CARDIOVASCULAR RISK STRATIFICATION OF PATIENTS WITH CHRONIC OBSTRUCTIVE PULMONARY DISEASE AND HEIGHTENED CARDIOVASCULAR RISK. The association between plasma high-sensitivity cardiac troponin I concentration and cardiovascular events in patients with chronic obstructive pulmonary disease and heightened cardiovascular risk was examined within the context of a double-blind randomised controlled trial of inhaled corticosteroids and bronchodilators (1 placebo arm and 3 different treatment arms). Plasma cardiac troponin I concentrations were measured with a high-sensitivity assay in a subgroup of 1,599 patients. The cardiovascular endpoint was a composite of cardiovascular death, myocardial infarction, stroke, unstable angina and transient ischaemic attack during follow-up of 1.5 years. Baseline plasma cardiac troponin I concentrations were above the lower limit of detection (1.0 ng/L) in 1,559 (97.5%) patients and were unaffected by inhaled therapies at 3 months (p > 0.05 for all). Compared with the lowest tertile (cardiac troponin I ≤3.0 ng/L), patients in the highest tertile (≥ 5.5 ng/L) were at greater risk of cardiovascular events (hazard ratio 3.0, 95% confidence interval 1.5 to 6.2, p=0.002) and cardiovascular death (hazard ratio 9.6, 95% confidence interval 2.6 to 35.6, p < 0.001) after adjustment for cardiovascular risk factors. There were no differences in COPD exacerbations between tertiles even after adjustment (p > 0.05). / REPRODUCIBILITY OF CORONARY 18F-FLUORIDE PET-CT IMAGING The inter-observer and scan-rescan reproducibility of coronary 18F-fluoride PET-CT imaging was investigated in 20 patients with clinically stable but high risk multi-vessel coronary artery disease who underwent repeated 18F-fluoride PET-CT scans 11.5±4.5 days apart. Scan analysis using the currently accepted approach of normalisation to a referent coronary segment (TBRREFERENT) identified 10 (50%) patients with evidence of focal coronary 18F-fluoride uptake and demonstrated moderate agreement across observers on a per-patient level (k = 0.56). This was similar to the level of agreement achieved with visual assessment alone (k = 0.64). Reproducibility was improved by semi-quantitative reporting combining visual assessment with a threshold uptake value for determining the presence of tracer uptake (k = 0.84). Using the optimised approach achieved excellent agreement on overall segmental uptake counts (intra-class correlation = 0.97). CONCLUSION: Cardiovascular diagnostic and prognostic assessments represent a complex endeavour and established tools for risk prediction can demonstrate suboptimal predictive accuracy when evaluated in patient cohorts that are independent of the population used for model derivation.

BACKGROUND Aortic stenosis is a common and potentially fatal condition in which fibro-calcific changes within the valve leaflets lead to the obstruction of blood flow. Severe symptomatic stenosis is an indication for aortic valve replacement and timely referral is essential to prevent adverse clinical events. Calcification is believed to represent the central process driving disease progression. 18F-Fluoride positron emission tomography computed tomography (PET-CT) and CT aortic valve calcium scoring (CT-AVC) quantify calcification activity and burden respectively. The overarching aim of this thesis was to evaluate the applications of these techniques to the study and management of aortic stenosis. METHODS AND RESULTS REPRODUCIBILITY The scan-rescan reproducibility of 18F-fluoride PET-CT and CT-AVC were investigated in 15 patients with mild, moderate and severe aortic stenosis who underwent repeated 18F-fluoride PET-CT scans 3.9±3.3 weeks apart. Modified techniques enhanced image quality and facilitated clear localization of calcification activity. Percentage error was reduced from ±63% to ±10% (tissue-to-background ratio most-diseased segment (MDS) mean of 1.55, bias -0.05, limits of agreement - 0·20 to +0·11). Excellent scan-rescan reproducibility was also observed for CT-AVC scoring (mean of differences 2% [limits of agreement, 16 to -12%]). AORTIC VALVE CALCIUM SCORE: SINGLE CENTRE STUDY Sex-specific CT-AVC thresholds (2065 in men and 1271 in women) have been proposed as a flow-independent technique for diagnosing severe aortic stenosis. In a prospective cohort study, the impact of CT-AVC scores upon echocardiographic measures of severity, disease progression and aortic valve replacement (AVR)/death were examined. Volunteers (20 controls, 20 with aortic sclerosis, 25 with mild, 33 with moderate and 23 with severe aortic stenosis) underwent CT-AVC and echocardiography at baseline and again at either 1 or 2-year time-points. Women required less calcification than men for the same degree of stenosis (p < 0.001). Baseline CT-AVC measurements appeared to provide the best prediction of subsequent disease progression. After adjustment for age, sex, peak aortic jet velocity (Vmax) ≥ 4m/s and aortic valve area (AVA) < 1 cm2, the published CT-AVC thresholds were the only independent predictor of AVR/death (hazard ratio = 6.39, 95% confidence intervals, 2.90-14.05, p < 0.001). AORTIC VALVE CALCIUM SCORE: MULTICENTRE STUDY CT-AVC thresholds were next examined in an international multicenter registry incorporating a wide range of patient populations, scanner vendors and analysis platforms. Eight centres contributed data from 918 patients (age 77±10, 60% male, Vmax 3.88±0.90 m/s) who had undergone ECG-gated CT within 3 months of echocardiography. Of these 708 (77%) had concordant echocardiographic assessments, in whom our own optimum sex-specific CT-AVC thresholds (women 1377, men 2062 AU) were nearly identical to those previously published. These thresholds provided excellent discrimination for severe stenosis (c-statistic: women 0.92, men 0.88) and independently predicted AVR and death after adjustment for age, sex, Vmax ≥4 m/s and AVA < 1 cm2 (hazards ratio, 3.02 [95% confidence intervals, 1.83-4.99], p < 0.001). In patients with discordant echocardiographic assessments (n=210), CT-AVC thresholds predicted an adverse prognosis. BICUSPID AORTIC VALVES Within the multicentre study, higher continuity-derived estimates of aortic valve area were observed in patients with bicuspid valves (n=68, 1.07±0.35 cm) compared to those with tri-leaflet valves (0.89±0.36 cm p < 0.001,). This was despite no differences in measurements of Vmax (p=0.152), or CT-AVC scores (p=0.313). The accuracy of AVA measurments in bicuspid valves was therefore tested against alternative markers of disease severity. AVA measurements in bicuspid valves demonstrated extremely weak associations with CT-AVC scores (r2=0.08, p=0.02) and failed to correlate with downstream markers of disease severity in the valve and myocardium and against clinical outcomes. AVA measurements in bicuspid patients also failed to independently predict AVR/death after adjustment for Vmax ≥4 m/s, age and gender. In this population CT-AVC thresholds (women 1377, men 2062 AU) again provided excellent discrimination for severe stenosis. CONCLUSIONS Optimised 18F-fluoride PET-CT scans quantify and localise calcification activity, consolidating its potential as a biomarker or end-point in clinical trials of novel therapies. CT calcium scoring of aortic valves is a reproducible technique, which provides diagnostic clarity in addition to powerful prediction of disease progression and adverse clinical events.

The results of evolution can be inferred from comparative studies of related organisms. In this doctoral thesis, I use phylogenetic comparative methods along side geometric morphometrics to analyse shape variation in order to infer evolution of the caecilian skull. Caecilians are elongate, limbless amphibians that superficially resemble snakes or earthworms, and use their head as a locomotory organ. I characterise large-scale patterns of cranial morphological diversity and quantify variation across the main family-level clades by describing patterns relating to phylogeny, disparity and ecology. Then I examine the origins and evolution of morphological variation in the skull by describing patterns relating to morphological integration and modularity. This thesis demonstrates a variety of existing statistical techniques that can be used to infer processes from large-scale evolutionary patterns in morphological data using non model organisms. Throughout the thesis, I show that the evolution of the caecilian skull to be multifaceted. On the patterns of diversity, the most striking is a "starburst" arrangement in shape space, which suggests that early in caecilian evolution ancestral lineages traversed greater expanses of the shape space, while subsequent phylogenetic divergence within the main clades entailed less morphological diversification. This may be related to early diversification into different ecological-niches, yet more data are needed to test this. The clades differ considerably in their cranial disparity, but there appears to be no unified pattern across the whole order that indicates disparity is coupled with clade age or speciation events. I show that aquatic species are more diverse than their terrestrial relatives, and that there is convergence of cranial shape among dedicated burrowers with eyes covered by bone. On the patterns of morphological integration and modularity, another remarkable finding is the caecilian cranium is modular with respect to two functional regions, the snout and the back of the cranium. Modularity is important for understanding the evolution of this structure. The main elements of the caecilian anterior skeleton, the cranium, mandible and atlas vertebra, reveal different patterns of morphological integration, suggesting different developmental and evolutionary processes are involved in sorting and maintaining new variation of each structure. Allometry is an important component of integration in each of the structures. Covariation of the cranium-mandible after size correction is significant and follows the same direction of shape change across all levels and as shown for allometry. In contrast, covariation of the cranium-atlas follows different directions at each level. These results suggest the two main joint of the caecilian skull differ substantially in their origin and evolution. I discuss the contribution made in this thesis to caecilian and evolutionary biology and offer an outlook of how theses findings can be used to initiate future studies to better understand of the evolution of the caecilian skull.

Coronary artery disease (CAD) is associated with significant morbidity and mortality. Coronary artery calcification (CAC) indicates presence of CAD. Absence of CAC is associated with very low risk of having CAD but not equal to zero. In this study, we aim at developing a clinical prediction tool to predict presence of obstructive CAD among patients with zero calcium score. We developed two models. A full prespecified model with 7 variables based on input from clinical experts, and a reduced model with 4 variables based on univariate screening. Both models showed an acceptable performance (c-statistics of 0.68 for both). Both models performed well when validated, externally for the full model and internally for the reduced one. We derived a clinical risk score of 20 points from the full model. We found that a score threshold of ≥ 14 is associated with presence of obstructive CAD with positive likelihood ratio of 5.5.