No description available.

No description available.

The image segmentation task divides an image into regions of similar pixels based on brightness, color, and texture, in which every pixel in the image is as- signed to a label. Segmentation is vital in numerous medical imaging applications, such as quantifying the size of tissues, the localization of diseases, treatment plan- ning, and surgery guidance. This thesis focuses on two medical image segmentation tasks: retinal vessel segmentation in fundus images and brain segmentation in 3D MRI images. Finally, we introduce LEON, a lightweight neural network for edge detection. The first part of this thesis proposes a lightweight neural network for retinal blood vessel segmentation. Our model achieves cutting-edge outcomes with fewer parameters. We obtained the most outstanding performance results on CHASEDB1 and DRIVE datasets with an F1 measure of 0.8351 and 0.8242, respectively. Our model has few parameters (0.34 million) compared to other networks such as ladder net with 1.5 million parameters and DCU-net with 1 million parameters. The second part of this thesis investigates the association between whole and re- gional volumetric alterations with increasing age in a large group of healthy subjects (n=6739, age range: 30–80). We used a deep learning model for brain segmentation for volumetric analysis to extract quantified whole and regional brain volumes in 95 classes. Segmentation methods are called edge or boundary-based methods based on finding abrupt changes and discontinuities in the intensity value. The third part of the thesis introduces a new Lightweight Edge Detection Network (LEON). The proposed approach is designed to integrate the advantages of the deformable unit and DepthWise Separable convolutions architecture to create a lightweight back- bone employed for efficient feature extraction. Our experiments on BSDS500 and NYUDv2 show that LEON, while requiring only 500000 parameters, outperforms the current lightweight edge detectors without using pre-trained weights. / Graduate / 2022-10-12

No / In this work, energy consumption in a middle vessel batch reactive distillation (MVBRD) column is considered for the production of lactic acid via hydrolysis of methyl lactate. A dynamic optimization problem incorporating a process model is formulated to minimize the batch time which consequently minimizes the total energy consumption. The problem is subject to constraints on the amount and purity of lactic acid. The optimisation variables are reflux ratio and/or reboil ratio which are treated as piecewise constant. The earlier work of the authors on energy consumption in conventional batch reactive distillation column (CBRD) for the same reaction system is used for comparative analysis with the energy consumption in MVBRD. As an example, for a given separation task, the optimization results show that MVBRD is capable of saving over 23 % energy compared to energy consumption in CBRD column for the same task.

This dissertation characterizes the geometry, kinematics, and physical properties of insect internal structures that make up the respiratory and circulatory systems. This characterization is necessary to better understand how these systems function to transport fluids at the microscale, and ultimately, how we might computationally model this flow. Chapter 2 describes the geometry of the insect tracheal system, specifically testing if Murray's law applies to this system using three-dimensional imaging of tracheal tubes. Chapter 3 begins to characterize the physical properties of insect hemolymph, specifically the viscosity and density of hemolymph, using experimental measurements. Because insects are strongly affected by environmental temperature, this chapter also explores how hemolymph viscosity may be affected by temperature. Chapter 4 builds on the results of Chapter 3, exploring the effects of developmental responses to temperature on hemolymph viscosity and properties, as well as performance of the insect using experimental measurements. Finally, Chapter 5 presents a kinematic and structural characterization of the insect heart using a variety of imaging techniques and analyses. / Doctor of Philosophy / Insect physiology and morphology has long been studied by biologists and entomologists, with many of the basic features understood and characterized. The insect circulatory and respiratory systems differ greatly from those of many other organisms. Physically, these systems transport fluids through microscale environments which include a variety of pumps, networks, and other structures that facilitate flow. Functionally, the circulatory and respiratory systems are largely decoupled, unlike in vertebrates. The respiratory system transports air directly to deliver oxygen to tissues, whereas the circulatory system transports various nutrients and other chemicals via hemolymph. With these unique differences, investigation of these major biological transport systems in insects is essential to fully understand their structure and function. This dissertation addresses many of the basic structural and physical properties of the insect respiratory and circulatory systems that are still unknown, despite growing engineering analysis. First, I measured specific geometric features of the insect tracheal network and determined if Murray’s law applies to this system. Second, I quantified the viscosity of insect hemolymph, including in response to temperature. To expand upon this relationship further, I measured hemolymph viscosity, hemolymph composition, and insect performance after temperature acclimation during development. Last, I investigated the morphology and kinematics of the insect heart, using many methods of imaging and analysis to measure structural features of the heart wall, including during function. Hemolymph properties and heart morphology provide the physical basis of flow production within the circulatory system. Understanding flow production within the circulatory system, as well as design features of the respiratory system, are crucial in the construction of mathematical models of both hemolymph and air flow within the insect.

This work focuses on the development and implementation of an autonomous path planning and obstacle avoidance algorithm for an autonomous surface vehicle (ASV) in a riverine environment. The algorithm effectively handles trap situations, which occur when the river bends away from the destination. In addition, the algorithm uses real-time sensor feedback to avoid obstacles. A general global route is proposed based on an a priori shoreline map. Then, local paths are calculated considering both the a priori data and measurements received from an obstacle sensor. These paths roughly follow the global path. The algorithm was tested on an ASV equipped with basic navigational sensors and an omnidirectional camera for obstacle detection, and experimentation verified its effectiveness. / Master of Science

Understanding the flow in stirred vessels can be useful for a wide number of industrial applications, like in mining, chemical and pharmaceutical processes. Remodeling and redesigning these processes may have a significant impact on the overall design characteristics, affecting directly product quality and maintenance costs. In most cases the flow around the rotating impeller blades interacting with stationary baffles can cause rapid changes of the flow characteristics, which lead to high levels of turbulence and higher shear rates. The flow is anisotropic and inhomogeneous over the entire volume. A better understanding and a detailed documentation of the turbulent flow field is needed in order to design stirred tanks that can meet the required operation conditions. This thesis describes efforts for accurate estimation of the velocity distribution and the turbulent characteristics (vorticity, turbulent kinetic energy, dissipation rate) in a cylindrical vessel agitated by a Rushton turbine (a disk with six flat blades) and in a tank typical of flotation cells. Results from simulations using FLUENT (a commercial CFD package) are compared with Time Resolved Digital Particle Image Velocimetry (DPIV) for baseline configurations in order to validate and verify the fidelity of the computations. Different turbulence models are used in this study in order to determine the most appropriate for the prediction of turbulent properties. Subsequently a parametric analysis of the flow characteristics as a function of the clearance height of the impeller from the vessel floor is performed for the Rushton tank as well as the flotation cell. Results are presented for both configurations along planes normal or parallel to the impeller axis, displaying velocity vector fields and contour plots of vorticity turbulent dissipation and others. Special attention is focused in the neighborhood of the impeller region and the radial jet generated there. This flow in this neighborhood involves even larger gradients and dissipation levels in tanks equipped with stators. The present results present useful information for the design of the stirring tanks and flotation cells, and provide some guidance on the use of the present tool in generating numerical solutions for such complex flow fields. / Master of Science

Introduction: Vascular injury is the crucial initiating event in atherosclerosis and is universal following percutaneous coronary intervention. The cellular response to this injury largely determines vessel outcome. Endothelial progenitor cells (EPCs) and their progeny, late outgrowth endothelial cells (EOCs) are thought to play an important role in this process and characterising this role would be valuable in better understanding vascular injury and repair. Methods: The radial artery in the context of transradial cardiac catheterisation was examined as a model of vascular injury with characterisation of structural injury, longitudinal function and EPC populations. To examine the role of late outgrowth endothelial cells a method for GMP-compliant cell culture and labelling with F18Fluorodeoxyglucose was developed with a view to conducting a cell-tracking study of human administration. Results: Radial artery function was reduced following transradial cardiac catheterisation with recovery over a period of three months. There was no correlation between recovery of arterial function and EPC populations as defined by conventional surface markers. A research grade protocol for EOC culture was successfully translated to a GMP-compliant process producing a viable, phenotypically homogeneous EOC product. Cells were successfully labelled with F18Fluorodeoxyglucose and whilst proliferation was reduced, acute viability and function were not compromised. Conclusion: The radial artery in the context of transradial cardiac catheterisation is a useful model of vascular injury and repair although recovery of vascular function does not appear to be influenced by EPC populations. GMP-compliant culture and labelling of EOCs is feasible and will allow examination of the physiology of these cells in vivo in man.

The accident at the Fukushima Daiichi nuclear power plant has shown us that there may be situations where the applied technology will not be able to successfully cool the reactor core. These situations may occur when more elements such as supply of energy to power the pumps and diesel generators are destroyed for example by tsunami or earthquake, or other not expected natural disasters. The inability of the residual heat removal leads to the melting of core, relocation to the bottom of reactor pressure vessel (RPV) and failure of RPV. Result of this accident may be containment failure and leakage of fission products into the environment. One way to prevent this scenario may be a passive system called IVR (In-Vessel Retention) by using external cooling of RPV that retains melted core in. This system counts with flooding of RPV´s shaft by water. After natural circulation of water provides the heat transfers from the wall of RPV. The applicability of IVR for VVER 1000 reactors is still in the course of research. However it´s already clear that the submersion of RPV shaft by water will not sufficient. Other elements as suitable insulation and RPV coating which provides a more intensive heat transfer from the walls of RPV will be needed.

<p>    Collateral connections in the brain, also known as Leptomeningeal Anastomoses, are connections between blood vessels originating from different arteries. Despite limited knowledge, they are suggested as an important contributor to cerebral stroke recovery that allows additional blood flow through the affected area. However, few databases and algorithms exist for this specific task of locating them. </p> <p>    In this paper, a MATLAB program is developed to find these connections and detect strokes to replace manual labeling by professionals.  The limited data available for this study are 23 2D microscopy images of mice cerebral vascular structures highlighted by dyes. In the images, strokes are shown to diminish the pixel count of vessels below 80\% compared to the healthy brain. Stroke classification error is greatly reduced by narrowing the scope from comparing the entire hemisphere to one smaller region.</p> <p>    A novel way of finding collateral connections is utilizing connected components. Connected components organize all adjacent pixels into a group. All collateral connections can be found on the border of two neighboring arterial flow regions, and belong to the same group of connected components with the arterial source from each side. </p> <p>    Along with finding collateral connections, a newly created coordinate system allows regions to be defined relative to the brain landmarks, based on the brain's center, orientation, and scale.</p> <p>    The method newly proposed in this paper combines stroke detection, brain coordinate system extraction, and collateral connection detection in stroke-affected mouse brains using only image processing techniques.  This allows a simpler, more explainable result on limited data than other techniques such as supervised machine learning.  In addition, the new method does not require ground truth and high image count for training. This automated process was successfully interpreted by medical experts, which allows for further research into automating collateral connection detection in 3D.</p>