No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

One way of reducing the examination time in magnetic resonance imaging (MRI) is to reduce the amount of raw data acquired, by performing so-called undersampling. Conventionally, MRI data is acquired line-by-line on a Cartesian grid. In the field of Cardiovascular Magnetic Resonance (CMR), however, radial k-space sampling is seen as a promising emerging technique for rapid image acquisitions, mainly due to its robustness against motion disturbances occurring from the beating heart. Whereas Cartesian undersampling will result in image aliasing, radial undersampling will introduce streak artifacts. The objective of this work was to train the deep learning architecture, CycleGAN, to reduce streak artifacts in radially undersampled CMR images, and to evaluate the model performance. A benefit of using CycleGAN over other deep learning techniques for this application is that it can be trained on unpaired data. In this work, CycleGAN network was trained on 3060 radial and 2775 Cartesian unpaired CMR images acquired in human subjects to learn a mapping between the two image domains. The model was evaluated in comparison to images reconstructed using another emerging technique called GRASP. Whereas more investigation is warranted, the results are promising, suggesting that CycleGAN could be a viable method for effective streak-reduction in clinical applications.

No description available.

Bridge inspection and infrastructure inspection are critical steps in the lifecycle of the built environment. Emerging technologies and data are driving factors which are disrupting the traditional processes for conducting these inspections. Because inspections are mainly conducted visually by human inspectors, this paper focuses on improving the visual inspection process with data-driven approaches. Data driven approaches, however, require significant data, which was sparse in the existing literature. Therefore, this research first examined the present state of the existing data in the research domain. We reviewed hundreds of image-based visual inspection papers which used machine learning to augment the inspection process and from this, we compiled a comprehensive catalog of over forty available datasets in the literature and identified promising, emerging techniques and trends in the field. Based on our findings in our review we contributed six significant datasets to target gaps in data in the field. The six datasets comprised of structural material segmentation, corrosion condition state segmentation, crack detection, structural detail detection, and bearing condition state classification. The contributed datasets used novel annotation guidelines and benefitted from a novel semi-automated annotation process for both object detection and pixel-level detection models. Using the data obtained from our collected sources, task-appropriate deep learning models were trained. From these datasets and models, we developed a change detection algorithm to monitor damage evolution between two inspection videos and trained a GAN-Inversion model which generated hyper-realistic synthetic bridge inspection image data and could forecast a future deterioration state of an existing bridge element. While the application of machine learning techniques in civil engineering is not wide-spread yet, this research provides impactful contribution which demonstrates the advantages that data driven sciences can provide to more economically and efficiently inspect structures, catalog deterioration, and forecast potential outcomes. / Doctor of Philosophy / Bridge inspection and infrastructure inspection are critical steps in the lifecycle of the built environment. Emerging technologies and data are driving factors which are disrupting the traditional processes for conducting these inspections. Because inspections are mainly conducted visually by human inspectors, this paper focuses on improving the visual inspection process with data-driven approaches. Data driven approaches, however, require significant data, which was sparse in the existing literature. Therefore, this research first examined the present state of the existing data in the research domain. We reviewed hundreds of image-based visual inspection papers which used machine learning to augment the inspection process and from this, we compiled a comprehensive catalog of over forty available datasets in the literature and identified promising, emerging techniques and trends in the field. Based on our findings in our review we contributed six significant datasets to target gaps in data in the field. The six datasets comprised of structural material detection, corrosion condition state identification, crack detection, structural detail detection, and bearing condition state classification. The contributed datasets used novel labeling guidelines and benefitted from a novel semi-automated labeling process for the artificial intelligence models. Using the data obtained from our collected sources, task-appropriate artificial intelligence models were trained. From these datasets and models, we developed a change detection algorithm to monitor damage evolution between two inspection videos and trained a generative model which generated hyper-realistic synthetic bridge inspection image data and could forecast a future deterioration state of an existing bridge element. While the application of machine learning techniques in civil engineering is not widespread yet, this research provides impactful contribution which demonstrates the advantages that data driven sciences can provide to more economically and efficiently inspect structures, catalog deterioration, and forecast potential outcomes.

Isolated inverter can provide galvanic isolation which is necessary for some applications with safety regulations. Traditionally, a two-stage configuration is widely applied with isolated dc-dc stage and a sinusoidal pulse-width-modulated (SPWM) dc-ac stage. However, this two-stage configuration suffers from more components count, more complex control and tend to have lower efficiency and lower power density. Meanwhile, a large dc bus capacitor is needed to attenuate the double line frequency from SPWM for two-stage configuration. Therefore, the single-stage approach including an isolated dc-rectified sine stage and a line frequency unfolder is preferable. Since the unfolder circuit is at line frequency being almost lossless, the isolated dc-rectified sine stage becomes critical. However, the relevant research for the single-stage isolated inverter is limited. People either utilize PWM based converter as dc-rectified sine stage with duty cycle adjustment or apply SRC or LLC resonant converter for better soft switching characteristics. For PWM based converter, hard switching restricts the overall inverter efficiency, while for SRC/LLC, enough wide voltage gain range and full range ZVS are the major issues. This dissertation aims to provide solutions for a high-efficiency, high-frequency resonant converter based single-stage soft-switching isolated inverter design. The LLC and LCLCL resonant converters are applied as the isolated dc-rectified sine stage with variable frequency modulation (VFM). Therefore, the rectified sine wave generation consists of many dc-dc conversion with different switching frequencies and an efficient dc-rectified sine stage design needs each dc-dc conversion to be with high efficiency. This dissertation will first propose the optimization methods for LLC converter dc-dc conversion. ZVS models are derived to ensure fully ZVS performance for primary side GaN devices. As a large part in loss breakdown, the optimization for transformer is essential. The LLC converter can achieve above 99% efficiency with proposed optimization approach. Moreover, the channel turn-off energy model is presented for a more accurate loss analysis. With all the design and optimization considerations, a MHz LLC converter based isolated inverter is designed and a hybrid modulation method is proposed, which includes full bridge (FB) VFM for output high line region and half bridge (HB) VFM for output low line region. By changing from FB to HB, the output voltage gain is reduced to half to have a wider voltage gain range. However, the total harmonic distortion (THD) of output voltage at light load will be impacted since the voltage gain will be higher with lighter load at the maximum switching frequency. A MHz LCLCL converter based isolated inverter is proposed for a better output voltage THD at light load conditions. The paralleled LC inside the LCLCL resonant tank can naturally create a zero voltage gain point at their resonant frequency, which shows superior performance for rectified sine wave generation. Besides the better THD performance, the LCLCL converter based isolated inverter also features for easier control, better ZVS performance and narrower switching frequency range. Meanwhile, the LCLCL based inverter topology has bi-directional power flow capability as well. With variable frequency modulation for ac-dc, this topology is still a single-stage solution compared to the traditional two-stage solution including PFC + LLC configuration. / Doctor of Philosophy / Inverters can convert dc voltage to ac voltage and typically people use two-stage approach with isolated dc-dc stage and dc-ac stage. However, this two-stage configuration suffers from more components count, more complex control and tend to have lower efficiency and lower power density. Therefore, the single-stage solution with dc-rectified sine wave stage and a line frequency unfolder becomes appealing. The unfolder circuit is to unfold the rectifier sine wave to an ac sine wave at the output. Since the unfolder is at line frequency and can be considered lossless, the key design is for the dc-rectified sine stage. The resonant converter featured for soft switching seems to be a good candidate. However, the inverter needs soft switching for the whole range and an enough wide voltage gain, which makes the design difficult, especially the target is high efficiency for the overall inverter. This dissertation aims to provide solutions for a high-efficiency, high-frequency resonant converter based single-stage soft-switching isolated inverter design. The LLC and LCLCL resonant converters are applied as the isolated dc-rectified sine stage with variable frequency modulation (VFM). Therefore, the rectified sine wave generation consists of many dc-dc conversion with different switching frequencies and an efficient dc-rectified sine stage design needs each dc-dc conversion to be with high efficiency. The design considerations and optimization methods for the LLC dc-dc conversion are firstly investigated. Based on these approaches, a MHz LLC converter based isolated inverter is designed with proposed hybrid modulation method. To further improve the light load performance, a MHz LCLCL converter based isolated inverter topology is proposed. The paralleled LC inside the LCLCL resonant tank can naturally create a zero voltage gain point which shows superior characteristics for rectified sine wave generation. Moreover, the LCLCL resonant converter based topology has bi-directional capability as well so it can work well for ac voltage to dc voltage conversion.