No description available.

No description available.

No description available.

No description available.

During the last few years, segmentation architectures based on deep learning achieved promising results. On the other hand, attention networks have been invented years back and used in different tasks but rarely used in medical applications. This thesis investigated four main attention mechanisms; Squeeze and Excitation, Dual Attention Network, Pyramid Attention Network, and Attention UNet to be used in medical image segmentation. Also, different hybrid architectures proposed by the author were tested. Methods were tested on a kidney tumor dataset and against UNet architecture as a baseline. One version of Squeeze and Excitation attention outperformed the baseline. Original Dual Attention Network and Pyramid Attention Network showed very poor performance, especially for the tumor class. Attention UNet architecture achieved close results to the baseline but not better. Two more hybrid architectures achieved better results than the baseline. The first is a modified version of Squeeze and Excitation attention. The second is a combination between Dual Attention Networks and UNet architecture. Proposed architectures outperformed the baseline by up to 3% in tumor Dice coefficient. The thesis also shows the difference between 2D architectures and their 3D counterparts. 3D architectures achieved more than 10% higher tumor Dice coefficient than 2D architectures.

This dissertation details the development and applications of two innovative types of optical filament-based impulsive Raman spectroscopy: filament-assisted Raman spectroscopy (FAIRS) and spectral-to-temporal amplitude mapping polarization spectroscopy (STAMPS). These techniques provide complimentary vibrational and rotational information on molecular systems of interest. Both are powerful due to their impulsive nature which allows for rapid measurement of entire Raman spectra. However, each type of spectroscopy utilizes the filament in a different manner. The recently reported vibrational technique, referred to as filament-assisted impulsive Raman spectroscopy, employs the pulse shortening and continuum generation of filamentation to impulsively excite a massive vibrational coherence in a molecular system for simultaneous measurement of all the Raman-active modes. In the first half of this dissertation, FAIRS is further developed and applied to a plethora of signature molecules. Radioactive decay signature molecules, including nitrogen oxides, ozone, and ions are detected via FAIR spectroscopy. Concurrent generation and detection of ozone, ionic, and excited-state molecules through filamentation is reported for the first time. Production of these species through the strong field chemistry of filamentation and their subsequent filament-driven excitation is a mark of sensitivity of FAIRS. Spatial studies of combustion species in a natural gas flame are also presented. FAIRS monitors the Raman signal intensities of known reactants and products as a function of vertical flame position. The appearance of combustion products as a function of flame height is also tracked. Spectral fringes overlapping the Raman-active modes are present in all measurements and enable more sensitive detection of low signal intensity species. The results described illustrate the potential of FAIRS for threat sensing applications. The rotational technique, referred to as spectral-to-temporal amplitude mapping polarization spectroscopy, temporally chirps the spectral content of the white-light continuum generated during filamentation to map the transient rotational rivals that are impulsively excited by a short pump pulse. In the second half of this dissertation, the initial development and testing, followed by the applications of STAMPS are described. STAMPS proves successful in mapping the rotational wavepacket rephasing of simple linear molecules, including nitrogen, oxygen, and carbon dioxide, as well as the more complicated asymmetric top molecules, ethylene and methanol. The application of STAMPS to the detection of nitrogen oxides and nitrous oxide, which are considered signatures of multiple threat substances and events, is demonstrated. A pressure study of nitrous oxide reveals dephasing effects as a function of time and pressure. These preliminary results also indicate the potential of STAMPS for hazard sensing applications. / Chemistry

The field of condensed matter computational physics has seen an explosion of applicability over the last 50+ years. Since the very first calculations with ENIAC and MANIAC the field has continued to pushed the boundaries of what is possible; from the first large-scale molecular dynamics simulation, to the implementation of Density Functional Theory and large scale Car-Parrinello molecular dynamics, to million-core turbulence calculations by Standford. These milestones represent not only technological advances but theoretical breakthroughs and algorithmic improvements as well. The work in this thesis was completed in the hopes of furthering such advancement, even by a small fraction. Here we will focus mainly on the calculation of electronic and structural properties of solids and liquids, where we shall implement a wide range of novel approaches that are both computational efficient and physically enlightening. To this end we routinely will work with maximally localized Wannier functions (MLWFs) which have recently seen a revival in mainstream scientific literature. MLWFs present us with interesting opportunity to calculate a localized orbital within the planewave formalism of atomistic simulations. Such a localization will prove to be invaluable in the construction of layer-based superlattice models, linear scaling hybrid functional schemes and model quasiparticle calculations. In the first application of MLWF we will look at modeling functional piezoelectricity in superlattices. Based on the locality principle of insulating superlattices, we apply the method of Wu et al to the piezoelectric strains of individual layers under iifixed displacement field. For a superlattice of arbitrary stacking sequence an accurate model is acquired for predicting piezoelectricity. By applying the model in the superlattices where ferroelectric and antiferrodistortive modes are in competition, functional piezoelectricity can be achieved. A strong nonlinear effect is observed and can be further engineered in the PbTiO 3 /SrTiO 3 superlattice and an interface enhancement of piezoelectricity is found in the BaTiO 3 /CaTiO 3 superlattice. The second project will look at The ionization potential distributions of hydrated hydroxide and hydronium which are computed within a many-body approach for electron excitations using configurations generated by ab initio molecular dynamics. The experimental features are well reproduced and found to be closely related to the molecular excitations. In the stable configurations, the ionization potential is mainly perturbed by solvent water molecules within the first solvation shell. On the other hand, electron excitation is delocalized on both proton receiving and donating complex during proton transfer, which shifts the excitation energies and broadens the spectra for both hydrated ions. The third project represents a work in progress, where we also make use of the previous electron excitation theory applied to ab initio x-ray emission spectroscopy. In this case we make use of a novel method to include the ultrafast core-hole electron dynamics present in such situations. At present we have shown only strong qualitative agreement with experiment. / Physics

Sliding is recognized as a dominant response mode for unanchored stocky components in nuclear facilities. Although unanchored components are themselves not safety-critical, their interaction with safety-critical systems and components during earthquake shaking can have significant consequences. It is therefore important to be able to accurately estimate the peak sliding displacement demands on unanchored components so that sufficient clearance is provided around them. In lieu of nonlinear time history analysis, the ASCE/SEI 43-05 standard provides an approximate method to estimate the maximum sliding displacement of sliding objects in nuclear facilities. The present paper assesses the procedure of the approximate method and compares its results to those of nonlinear time history analysis. The study finds that the ASCE 43-05 approximate method provides conservative sliding estimates overall and that is based on the three components of 7 modified and 159 real earthquake motions used in this study. It is concluded that the ASCE 43-05 approximate method offers reasonable sliding estimates of components in nuclear facilities. / Thesis / Master of Applied Science (MASc)

<p> In nuclear astrophysics, there are still undiscovered areas involving unstable nuclei, like the nucleosynthesis in exploding stars. The unstable nuclei are extremely difficult to produce in the laboratory for study since they decay away quickly once they are formed. To make the unstable nuclei available in the laboratory, Canada has built one of the best facilities in the world for nuclear astrophysics studies with radioactive beams, called TRIUMF-ISAC. Its recent upgrade to ISAC-2 can produce even heavier radioactive beams with higher energy. To fully utilize the ISAC-2 facility, a high-segmented HPGe γ-ray detector-TIGRESS, has been developed to study exotic nuclei using the well-established technique of Coulomb Excitation.</p> <p> As an essential part of the TIGRESS facility, auxiliary detectors, such as silicon detectors, are used to improve the sensitivity of the experiments through Doppler correction. Two types of silicon detectors will be used. One is the CD-shaped double-sided silicon strip detector, which is the topic of this thesis. The other one is the silicon barrel detector. This thesis will discuss the computer simulations and experimental tests of the CD detector.</p> <p> Simulations were made using a simulation toolkit GEANT4, which was run under the Linux system. The goal was to test the hypothesis that measurements of the energies of heavy ions in Coulomb excitation may improve the Doppler correction by constraining the scattering location in a thick target. To test the performance of the simulated CD detector, simple simulations were performed for shooting calibration alpha particles from an 241Am source first directly onto the CD and second through a thin target onto the CD. The goal of additional simulations was to attempt to determine the corresponding scattering depth in a thick target by putting an energy cut on the spectrum of the scattered beam detected in the CD. These simulation results verify the possibility of such localization, though the resolution is not very good.</p> <p> Experimental tests for the CD detector are aimed to obtain the energy resolution for each strip element of the CD and to have at least 8 channels instrumented in our test lab before this detector is sent to TIGRESS facility. The test station including electronics and detector systems and all results from initial tests are described in detail in this thesis.</p> / Thesis / Master of Science (MSc)

The author proposes a novel functional regression method for parameter estimation and adaptive control in this dissertation. In the functional regression method, the regressors and a signal which contains the information of the unknown parameters are either determined from raw measurements or calculated as the functions of the measurements. The novel feature of the method is that the algorithm maps the regressors to the functionals which are represented in terms of customized test functions. The functionals are updated continuously by the evolution laws, and only an infinite number of variables are needed to compute the functionals. These functionals are organized as the entries of a matrix, and the parameter estimates are obtained using either the generalized inverse method or the transpose method. It is shown that the schemes of some conventional adaptive methods are recaptured if certain test function designs are employed. It is proved that the functional regression method guarantees asymptotic convergence of the parameter estimation error to the origin, if the system is persistently excited. More importantly, in contrast to the conventional schemes, the parameter estimation error may be expected to converge to the origin even when the system is not persistently excited. The novel adaptive method are also applied to the Model Reference Adaptive Controller (MRAC) and adaptive observer. It is shown that the functional regression method ensures asymptotic stability of the closed loop systems. Additionally, the studies indicate that the transient performance of the closed loop systems is improved compared to that of the schemes using the conventional adaptive methods. Besides, it is possible to analyze the transient responses a priori of the closed loop systems with the functional regression method. The simulations verify the theoretical analyses and exhibit the improved transient and steady state performances of the closed loop systems. / Ph. D.