Turbulent blood flow is involved in the pathogenesis of several cardiovascular diseases. While it is known that turbulence is present in patients with obstructive disease in the major vessels, the magnitude and impact of turbulence in the normal heart and aorta is still relatively unexplored. Besides, existing analysis method of the blood flow is a labour intensive process and requires excessive amount of time. A method to automatically create hemodynamic atlases has been developed, using 4D Flow magnetic resonance imaging (MRI), a powerful tool to measure blood flow characteristics. The resulting atlases show the expected blood flow characteristics in the aorta for a group of similar subjects. Application of the method in healthy young and healthy old has shown significant differences in kinetic energy and turbulent kinetic energy in the aortic flow.

Microscopy image processing is an emerging and quickly growing field in medical imaging research area. Recent advancements in technology including higher computation power, larger and cheaper storage modules, and more efficient and faster data acquisition devices such as whole-slide imaging scanners contributed to the recent microscopy image processing research advancement. Most of the methods in this research area either focus on automatically process images and make it easier for pathologists to direct their focus on the important regions in the image, or they aim to automate the whole job of experts including processing and classifying images or tissues that leads to disease diagnosis. This dissertation is consisted of four different frameworks to process microscopy images. All of them include methods for segmentation either as the whole suggested framework or the initial part of the framework for future feature extraction and classification. Specifically, the first proposed framework is a general segmentation method that works on histology images from different tissues and segments relatively solid nuclei in the image, and the next three frameworks work on cervical microscopy images, segmenting cervical nuclei/cells. Two of these frameworks focus on cervical tissue segmentation and classification using histology images and the last framework is a comprehensive segmentation framework that segments overlapping cervical cells in cervical cytology Pap smear images. One of the several commonalities among these frameworks is that they all work at the region level and use different region features to segment regions and later either expand, split or refine the segmented regions to produce the final segmentation output. Moreover, all proposed frameworks work relatively much faster than other methods on the same datasets. Finally, proving ground truth for datasets to be used in the training phase of microscopy image processing algorithms is relatively time-consuming, complicated and costly. Therefore, I designed the frameworks in such a way that they set most (if not all) of the parameters adaptively based on each image that is being processed at the time. All of the included frameworks either do not depend on training datasets at all (first three of the four discussed frameworks) or need very small training datasets to learn or set a few parameters.

This study discusses reject film analyses (RFAs) before and after the implementation of a quality improvement intervention. RFAs were undertaken to investigate the effect of the introduction and use of exposure charts (ECs) on department and student reject rates of extremity radiographs. Methods: A quantitative comparative pre and post-treatment research design was used. Data was collected from the x-ray departments of two training hospitals in Windhoek, Namibia over a five month period. A retrospective RFA was conducted to determine the department and student reject rates for both departments before intervention. Emphasis was placed on exposure related reject films. ECs were compiled and introduced at Katutura State Hospital (venue B) by the researcher. The students were instructed to use these charts. At Windhoek Central Hospital (venue A) no ECs were used. A prospective RFA was conducted to establish department and student reject rates at both hospitals after the intervention at venue B. Results: During the retrospective phase the department reject rate for venue A was 21 percent while the student reject rate was 23 percent. At venue B 24 percent and 26 percent were scored respectively. Students at venue A produced rejected radiographs due to overexposure (49 percent) and underexposure (23 percent), whilst 37 percent was recorded for both causes at venue B. At venue A, 35 percent of films were rejected due to incorrect mAs selection, at venue B the figure was 42 percent. Undiagnostic radiographs due to inaccurate kV selection comprised 62 percent for venue A and 59 percent for venue B. During the prospective phase the department reject rate for venue A was 20 percent and that of the students was 19 percent. For venue B 12 percent and 11 percent were scored respectively. At venue A radiographs rejected due to over and underexposure were 43 percent and 33 percent respectively while those at venue B were 33 percent and 34 percent. Incorrect mAs selection caused 33 percent of discarded films at venue A and 38 percent at venue B. The figures for inaccurate kV selection were 68 percent and 62 percent for venues A and B. Conclusions: The introduction and use of ECs lowered the student reject rate at venue B in the prospective phase.

This thesis investigates if support vector machine classification is a suitable approach when performing automatic segmentation of knee cartilage using quantitative magnetic resonance imaging data. The data sets used are part of a clinical project that investigates if patients that have suffered recent knee damage will develop cartilage damage. Therefore the thesis also investigates if the segmentation results can be used to predict the clinical outcome of the patients. Two methods that perform the segmentation using support vector machine classification are implemented and evaluated. The evaluation indicates that it is a good approach for the task, but the implemented methods needs to be further improved and tested on more data sets before clinical use. It was not possible to relate the cartilage properties to clinical outcome using the segmentation results. However, the investigation demonstrated good promise of how the segmentation results, if they are improved, can be used in combination with quantitative magnetic resonance imaging data to analyze how the cartilage properties change over time or vary between knees.

abstract: Recently, a well-designed and well-trained neural network can yield state-of-the-art results across many domains, including data mining, computer vision, and medical image analysis. But progress has been limited for tasks where labels are difficult or impossible to obtain. This reliance on exhaustive labeling is a critical limitation in the rapid deployment of neural networks. Besides, the current research scales poorly to a large number of unseen concepts and is passively spoon-fed with data and supervision. To overcome the above data scarcity and generalization issues, in my dissertation, I first propose two unsupervised conventional machine learning algorithms, hyperbolic stochastic coding, and multi-resemble multi-target low-rank coding, to solve the incomplete data and missing label problem. I further introduce a deep multi-domain adaptation network to leverage the power of deep learning by transferring the rich knowledge from a large-amount labeled source dataset. I also invent a novel time-sequence dynamically hierarchical network that adaptively simplifies the network to cope with the scarce data. To learn a large number of unseen concepts, lifelong machine learning enjoys many advantages, including abstracting knowledge from prior learning and using the experience to help future learning, regardless of how much data is currently available. Incorporating this capability and making it versatile, I propose deep multi-task weight consolidation to accumulate knowledge continuously and significantly reduce data requirements in a variety of domains. Inspired by the recent breakthroughs in automatically learning suitable neural network architectures (AutoML), I develop a nonexpansive AutoML framework to train an online model without the abundance of labeled data. This work automatically expands the network to increase model capability when necessary, then compresses the model to maintain the model efficiency. In my current ongoing work, I propose an alternative method of supervised learning that does not require direct labels. This could utilize various supervision from an image/object as a target value for supervising the target tasks without labels, and it turns out to be surprisingly effective. The proposed method only requires few-shot labeled data to train, and can self-supervised learn the information it needs and generalize to datasets not seen during training. / Dissertation/Thesis / Doctoral Dissertation Computer Science 2020

Statistical modelling has become a ubiquitous tool for analysing of morphological variation of bone structures in medical images. For radiological images, the shape, relative pose between the bone structures and the intensity distribution are key features often modelled separately. A wide range of research has reported methods that incorporate these features as priors for machine learning purposes. Statistical shape, appearance (intensity profile in images) and pose models are popular priors to explain variability across a sample population of rigid structures. However, a principled and robust way to combine shape, pose and intensity features has been elusive for four main reasons: 1) heterogeneity of the data (data with linear and non-linear natural variation across features); 2) sub-optimal representation of three-dimensional Euclidean motion; 3) artificial discretization of the models; and 4) lack of an efficient transfer learning process to project observations into the latent space. This work proposes a novel statistical modelling framework for multiple bone structures. The framework provides a latent space embedding shape, pose and intensity in a continuous domain allowing for new approaches to skeletal joint analysis from medical images. First, a robust registration method for multi-volumetric shapes is described. Both sampling and parametric based registration algorithms are proposed, which allow the establishment of dense correspondence across volumetric shapes (such as tetrahedral meshes) while preserving the spatial relationship between them. Next, the framework for developing statistical shape-kinematics models from in-correspondence multi-volumetric shapes embedding image intensity distribution, is presented. The framework incorporates principal geodesic analysis and a non-linear metric for modelling the spatial orientation of the structures. More importantly, as all the features are in a joint statistical space and in a continuous domain; this permits on-demand marginalisation to a region or feature of interest without training separate models. Thereafter, an automated prediction of the structures in images is facilitated by a model-fitting method leveraging the models as priors in a Markov chain Monte Carlo approach. The framework is validated using controlled experimental data and the results demonstrate superior performance in comparison with state-of-the-art methods. Finally, the application of the framework for analysing computed tomography images is presented. The analyses include estimation of shape, kinematic and intensity profiles of bone structures in the shoulder and hip joints. For both these datasets, the framework is demonstrated for segmentation, registration and reconstruction, including the recovery of patient-specific intensity profile. The presented framework realises a new paradigm in modelling multi-object shape structures, allowing for probabilistic modelling of not only shape, but also relative pose and intensity as well as the correlations that exist between them. Future work will aim to optimise the framework for clinical use in medical image analysis.

Image registration is the transformation of different sets of images into one coordinate system in order to align and overlay multiple images. Image registration is used in many fields such as medical imaging, remote sensing, and computer vision. It is very important in medical research, where multiple images are acquired from different sensors at various points in time. This allows doctors to monitor the effects of treatments on patients in a certain region of interest over time. In this thesis, artificial neural networks with curvelet keypoints are used to estimate the parameters of registration. Simulations show that the curvelet keypoints provide more accurate results than using the Discrete Cosine Transform (DCT) coefficients and Scale Invariant Feature Transform (SIFT) keypoints on rotation and scale parameter estimation.

The purpose is trying to classify the Liver Fibrosis stage using Gadoxetic Acid-EnhancedMR Images.  In the very beginning, a method proposed by one Korean group is being examined and trying to reproduce their result. However, the performance is not as impressive as theirs. Then, some gray-scale image feature extraction methods are used. Last but not least, the hottest method in recent years - ConvolutionNeural Network(CNN) was utilized. Finally, the performance has been evaluated in both methods. The result shows that with manual feature extraction, the Adaboost model works pretty well that AUC achieves 0.9. Besides, the AUC of ResNet-18 network - a deep learning architecture, can reach 0.93. Also, all the hyperparameters and training settings used on ResNet-18 can be transferred to ResNet-50/ResNet-101/InceptionV3 very well. The best model that can be obtained is ResNet-101which has an AUC of 0.96 - higher than all current publications for machine learning methods for staging liver fibrosis.

DIRA-3D is an iterative model-based reconstruction method for dual-energy helical CT whose goal is to determine the material composition of the patient from accurate linear attenuation coefficients (LACs). Possible applications are, for example, to aid in calculations of radiation transport and dose calculations in brachytherapy with low energy photons, and in proton therapy. There was a need to replace the current image reconstruction method, the PI-method, with a weighted filtered backprojection (wFBP) algorithm for image reconstruction, since wFBP is used for image reconstruction in Siemens's CT-scanners. The new DIRA-3D algorithm implemented the program take for cone-beam projection generation and the FreeCT wFBP algorithm for image reconstruction. Experiments showed that the accuracies of the resulting LACs for the DIRA-3D algorithm using wFBP for image reconstruction were comparable to the one using the PI-method for image reconstruction. The relative LAC errors reached a value below 0.2% after 10 iterations.

Each year million of babies are born pre-term, some of these pre-term births occur due to the motherhaving a too soft cervix which can not withstand the forces the baby exposes it to. The aim of thisstudy was to implement and evaluate a programmable shear wave elastography ultrasound system forcervical applications and investigate the optimal settings of shear wave elastography push voltage andshear wave elastography push focus depth. Shear wave elastography is an ultrasound based imagingmodality aiming to evaluate the tissue elasticity by using acoustic radiation forces to induce shear waves.The propagation of the shear waves through the tissue is then tracked in order to calculate the shearwave velocity which is related to the tissue elasticity. B-mode imaging, pushing sequence and planewave imaging have been implemented and measurements have been conducted on four cervical polyvinylalcohol phantoms. The acquired data has been post-processed using Loupas 2D-autocorrector to gainthe axial displacement and enabling tracking of the shear waves to allow evaluation and optimizationof the implemented method. The implemented shear wave technique showed to be able to distinguishcervical phantoms of dierent elasticity and a high pushing voltage and shallow focus push depth havebeen found to produce the most reliable results.