Search results

Title: RADIAL MAP ASSESSMENT APPROACH FOR DEEP LEARNING DENOISED CARDIAC MAGNETIC RESONANCE RECONSTRUCTION SHARPNESS
Creator: Mo, Fei
Date: 2021
Description: Deep Learning (DL) and Artificial Intelligence (AI) play important roles in the computer-aided medical diagnostics and precision medicine...
Show moreDeep Learning (DL) and Artificial Intelligence (AI) play important roles in the computer-aided medical diagnostics and precision medicine fields, capable of complementing human operators in disease diagnosis and treatment but optimizing and streamlining medical image display. While incredibly powerful, images produced via Deep Learning or Artificial Intelligence should be analyzed critically in order to be cognizant of how the algorithms are producing the new image and what the new imagine is. One such opportunity arose in the form of a unique collaborative project: the technical development of an image assessment tool that would analyze outputs between DL-based and non DL-based Magnetic Resonance Imaging reconstruction methods.More specifically, we examine the operator input dependence of the existing reference method in terms of accuracy and precision performance, and subsequently propose a new metric approach that preserves the heuristics of the intended quantification, overcomes operator dependence, and provides a relative comparative scoring approach that may normalize for angular dependence of examined images. In chapter 2 of this thesis, we provide a background description pertaining to the two imaging science principles that yielded our proposed method description and study design. First, if treated naively, the examined linear measurement approach exhibits potential bias with respect to the coordinate lattice space of the examined image. Second, the examined DL-based image reconstruction methods used in this thesis warrants an elaborate and explicit description of the measured noise and signal present in the reconstructed images. This specific reconstruction approach employs an iterative scheme with an embedded DL-based substep or filter to which we are blinded. In chapters 3 and 4 of this thesis, the imaging and DL-based image reconstruction experiments are described. These experiments employ cardiac MRI datasets from multiple clinical centers. We first outline the clinical and technical background for this approach, and then examine the quality of DL-based reconstructed image sharpness by two alternative methods: 1) by employing the gold-standard method that addresses the lattice point irregularity using a ‘re-gridding’ method, and 2) by applying our novel proposed method inspired by radial MRI k-space sampling, which exploits the mathematical properties of uniform radial sampling to yield the target voxel counts in the ‘gridded’ polar coordinate system. This new measure of voxel counts is shown to overcome the limitation due to the operator-dependence for the conventional approach. Furthermore, we propose this metric as a relative and comparative index between two alternative reconstruction methods from the same MRI k-space.
Show less

Title: DEEP LEARNING AND COMPUTER VISION FOR INDUSTRIAL APPLICATIONS: CELLULAR MICROSCOPIC IMAGE ANALYSIS AND ULTRASOUND NONDESTRUCTIVE TESTING
Creator: Yuan, Yu
Date: 2022
Description: For decades, researchers have sought to develop artificial intelligence (AI) systems that can help human beings on decision making, data...
Show moreFor decades, researchers have sought to develop artificial intelligence (AI) systems that can help human beings on decision making, data analysis and pattern recognition applications where analytical methods are ineffective. In recent years, Deep Learning (DL) has been proven to be an effective AI technique that can outperform other methods in applications such as computer vision, natural language processing, autonomous driving. Realizing the potential of deep learning techniques, researchers have also started to apply deep learning on other industrial applications. Today, deep learning based models are used to innovate and accelerate automation, guidance, and decision making in various industries including automotive industry, pharmaceutical industry, finance, agriculture and more. In this research, several important industrial applications (on Biomedicine and Non-Destructive Testing) utilizing deep learning algorithms will be introduced and analyzed. The first biopharmaceutical application focuses on developing a deep learning based model to automate the visual inspection process in Median Tissue Culture Infectious Dose(TCID50). TCID50 is one of the most popular methods for viral quantification. An important step of TCID50 is to visually inspect the sample and decide if it exhibits cytopathic effect(CPE) or not. Two novel models have been developed to detect CPE in microscopic images of cell culture in 96 well-plates. The first model consists of a convolutional neural network (CNN) and support vector machine(SVM). The second model is a fully convolutional network (FCN) followed by morphological post-processing steps. The models are tested on 4 cell lines and achieve very high accuracy. Another biopharmaceutical application developed for cellular microscopic images is the clonal selection. Clonal selection is one of the mandatory steps in cell line development process. It focuses on verifying the clonality of the cell culture. The researchers used to visually inspect the microscopic images to verify the clonality. In this work, a novel deep learning based model and a workflow is developed to accelerate the process. This algorithm consists of multiple steps, including image analysis after incubation to detect the cell colonies, and verify its clonality in day0 image. The results and common mis-classification cases are shown in this thesis. Image analysis method is not the only technology that has been advancing for cellular image analysis in biopharmaceutical industry. A new class of instruments are currently used in biopharmaceutical industry which enable more opportunities for image analysis. To make the most of these new instruments, a convolutional neural network based architecture is used to perform accurate cell counting and cell morphology based segmentation. This analysis can provide more insight of the cells at very early stage in characterization process of cell line development. The architecture and the testing results are presented in this work. The proposed algorithm has achieved very high accuracy on both applications, and the cell morphology based segmentation enables a brand new feature for scientists to predict the potential productivity of the cells. Next part of this dissertation is focused on hardware implementation of Ultrasonic Non-Destructive Testing (NDT) methods based on deep learning, which can be highly useful in flaw detection and classification applications. With the help of a smart and mobile Non-Destructive Testing device, engineers can accurately detect and locate the flaws inside the materials without reliance on high performance computation resources. The first NDT application presents a hardware implementation of a deep learning algorithm on Field-programmable gate array(FPGA) for Ultrasound flaw detection. The Ultrasound flaw detection algorithm consists of a wavelet transform followed by a LeNet inspired convolutional neural network called Ultra-LeNet. This work is focused on implementing the computationally difficult part of this algorithm: Ultra-LeNet, so that it can be used in the field where high performance computation resources (e.g., AWS) are not accessible. The implementation uses resource partitioning to design two dedicated pipelined accelerators for convolutional layers and fully connected layers respectively. Both accelerators utilize loop unrolling, loop pipelining and batch processing techniques to maximize the throughput. The comparison to other work has shown that the implementation has achieved higher hardware utilization efficiency. The second NDT application is also focused on implementing a deep learning based algorithm for Ultrasound flaw detection on a FPGA. Instead of implementing the Ultra-LeNet, the deep learning model used in this application is Meta-learning based Siamese Network, which is capable for multi-class classification and it can also classify a new class even if it does not appear in the training dataset with the help of automated learning features. The hardware implementation is significantly different than the previous algorithm. In order to improve the inference operation efficiency, the model is compressed with both pruning and quantization, and the FPGA implementation is specifically designed to accelerate the compressed CNN with high efficiency. The CNN model compression method and hardware design are novel methods introduced in this work. Comparison against other compressed CNN accelerators is also presented.
Show less

Title: Prediction and Control of In-Cylinder Processes in Heavy-Duty Engines Using Alternative Fuels
Creator: Pulpeiro Gonzalez, Jorge
Date: 2024
Description: This Ph.D. thesis focuses on advancing diagnostic techniques and control-oriented models to enhance the efficiency and performance of internal...
Show moreThis Ph.D. thesis focuses on advancing diagnostic techniques and control-oriented models to enhance the efficiency and performance of internal combustion (IC) engines, particularly heavy-duty engines utilizing alternative fuels. The research endeavors to contribute to the field of model-based control of engines through the development and implementation of innovative methodologies. The primary emphasis is on the development of diagnostic methods, control-oriented models and advanced control strategies for compression ignition engines using alternative fuels. The first key topic explores the determination of the Most Representative Cycle for Combustion Phasing Estimation based on cylinder pressure measurements. The method developed extracts crucial information from experimental data obtained from four distinct engines: the heavy-duty single-cylinder GCI engine, the light-duty multi-cylinder diesel engine, a CFR engine, and a single-cylinder light-duty Spark Ignition (SI) engine. This work lays the foundation for precise combustion phasing estimation, a critical parameter for engine control. The second major contribution involves the development of control-oriented models for Variable Geometry Turbochargers (VGT) and inter-coolers. Two models are established: a data-driven turbocharger model and an empirical inter-cooler model. These models are meticulously calibrated and validated using experimental data from a multi-cylinder light-duty diesel engine, providing valuable insights into the behavior of these components under varying conditions. The outcomes contribute to facilitate predictive control of engine air systems. The third core aspect of the thesis revolves around Model Predictive Control of Combustion Phasing in heavy-duty compression-ignition engines utilizing alternative fuels. A combustion phasing and engine load model is derived from experimental data and incorporated into an MPC framework. The MPC strategy is subsequently tested in the heavy-duty GCI test cell and compared against a conventional Proportional-Integral-Derivative (PID) control strategy. The results showcase the effectiveness of the MPC approach in achieving precise control of combustion phasing, demonstrating its potential for optimizing engine performance. In summary, this Ph.D. thesis contributes significantly to the field of engine controls by advancing diagnostic techniques, control-oriented models, and implementing a cutting-edge MPC-based control strategy for compression ignition engines using alternative fuels. The research findings not only enhance the understanding of in-cylinder processes but also pave the way for more efficient and sustainable heavy-duty engines using alternative fuels.
Show less

Title: High-latitude plasma drift structuring from a first principles ionospheric model
Creator: Kim, Heejin
Date: 2020
Description: In the high-latitude ionosphere dense plasma formations called polar cap patches are sometimes observed. These patches are often associated...
Show moreIn the high-latitude ionosphere dense plasma formations called polar cap patches are sometimes observed. These patches are often associated with ionospheric scintillation, a rapid fluctuation in the amplitude and phase of a radio signal that degrades communications and navigation systems. Predicting polar cap patch movement across the polar cap is an important subject for enabling forecasting of the scintillation.Lagrangian coherent structures (LCSs) are ridges indicating regions of maximum fluid separation in a time-varying flow. In previous studies, the Ionosphere-Thermosphere Algorithm for Lagrangian Coherent Structures (ITALCS) predicted the location of LCSs. These LCSs were shown to constrain polar cap patch source and transport regions for flow assumed to due to $\vec{E} \times \vec{B}$ plasma drift. The LCSs were predicted based on an empirical model of the high-latitude electric field for $\vec{E}$. In this thesis, the LCSs are generated using the first principles ionospheric model SAMI3 (SAMI3 is Another Model of the Ionosphere) as the model for electric field. The work relies on an understanding of various magnetic coordinate systems in space science, and includes three different approaches for attempting to generate the $\vec{E} \times \vec{B}$ drift as the flow fields that are to input to ITALCS. Finally, a representative LCS result is obtained with SAMI3 and shown to be at the high latitudes on the dayside, similar to prior work, but spanning a shorter longitudinal range.
Show less

Title: Estimation of Platinum Oxide Degradation in Proton Exchange Membrane Fuel Cells
Creator: Ahmed, Niyaz Afnan
Date: 2024
Description: The performance and durability of Proton Exchange Membrane Fuel Cells (PEMFCs) can be significantly hampered due to the degradation of the...
Show moreThe performance and durability of Proton Exchange Membrane Fuel Cells (PEMFCs) can be significantly hampered due to the degradation of the platinum catalyst. The production of platinum oxide is a major cause of the degradation of the fuel cell system, negatively affecting its performance and durability. In order to predict and prevent this degradation, this research examines a novel method to estimate degradation due to platinum oxide formation and predict the level of platinum oxide coverage over time. Mechanisms of platinum oxide formation are outlined and two methods are compared for platinum oxide estimation. Linear regression and two Artificial Neural Network (ANN) models, including a Recurrent Neural Network (RNN) and Feed-forward Back Propagation Neural Network (FFBPNN), are compared for estimation. The estimation model takes into account the influence of cell temperature and relative humidity.Evaluation of relative errors (RE) and root mean square error (RMSE) illustrates the superior performance of RNN in contrast to GT-Suite and FFBPNN. However, both RNN and GT-Suite showcase an average error rate below 5% while the FFBPNN had a higher error rate of approximately 7%. The RMSE of RNN shows mostly less compared to FFBPNN and GT-Suite, however, at 50% training data, GT-Suite shows lowest RMSE. These findings indicate that GT-Suite can be a valuable tool for estimating platinum oxide in fuel cells with a relatively low RE, but the RNN model may be more suitable for real-time estimation of platinum oxide degradation in PEM fuel cells, due to its accurate predictions and shorter computational time. This comprehensive approach provides crucial insights for optimizing fuel cell efficiency and implementing effective maintenance strategies.
Show less

Title: Prediction and Control of In-Cylinder Processes in Heavy-Duty Engines Using Alternative Fuels
Creator: Pulpeiro Gonzalez, Jorge
Date: 2024
Description: This Ph.D. thesis focuses on advancing diagnostic techniques and control-oriented models to enhance the efficiency and performance of internal...
Show moreThis Ph.D. thesis focuses on advancing diagnostic techniques and control-oriented models to enhance the efficiency and performance of internal combustion (IC) engines, particularly heavy-duty engines utilizing alternative fuels. The research endeavors to contribute to the field of model-based control of engines through the development and implementation of innovative methodologies. The primary emphasis is on the development of diagnostic methods, control-oriented models and advanced control strategies for compression ignition engines using alternative fuels. The first key topic explores the determination of the Most Representative Cycle for Combustion Phasing Estimation based on cylinder pressure measurements. The method developed extracts crucial information from experimental data obtained from four distinct engines: the heavy-duty single-cylinder GCI engine, the light-duty multi-cylinder diesel engine, a CFR engine, and a single-cylinder light-duty Spark Ignition (SI) engine. This work lays the foundation for precise combustion phasing estimation, a critical parameter for engine control. The second major contribution involves the development of control-oriented models for Variable Geometry Turbochargers (VGT) and inter-coolers. Two models are established: a data-driven turbocharger model and an empirical inter-cooler model. These models are meticulously calibrated and validated using experimental data from a multi-cylinder light-duty diesel engine, providing valuable insights into the behavior of these components under varying conditions. The outcomes contribute to facilitate predictive control of engine air systems. The third core aspect of the thesis revolves around Model Predictive Control of Combustion Phasing in heavy-duty compression-ignition engines utilizing alternative fuels. A combustion phasing and engine load model is derived from experimental data and incorporated into an MPC framework. The MPC strategy is subsequently tested in the heavy-duty GCI test cell and compared against a conventional Proportional-Integral-Derivative (PID) control strategy. The results showcase the effectiveness of the MPC approach in achieving precise control of combustion phasing, demonstrating its potential for optimizing engine performance. In summary, this Ph.D. thesis contributes significantly to the field of engine controls by advancing diagnostic techniques, control-oriented models, and implementing a cutting-edge MPC-based control strategy for compression ignition engines using alternative fuels. The research findings not only enhance the understanding of in-cylinder processes but also pave the way for more efficient and sustainable heavy-duty engines using alternative fuels.
Show less

Title: Development of data assimilation for analysis of ion drifts during geomagnetic storms
Creator: Hu, Jiahui
Date: 2024
Description: The primary objective of this dissertation is to gain insight into geomagnetic storm effects at mid-latitudes induced by solar activity....
Show moreThe primary objective of this dissertation is to gain insight into geomagnetic storm effects at mid-latitudes induced by solar activity. Geomagnetic storms affect our everyday lives because they give rise to transient signal loss, data transmission errors, negatively impacting users of satellite navigation systems. The Nighttime Localized Ionospheric Enhancement (NILE) is a localized plasma enhancement that because it is not well understood, drives the design of satellite-based augmentationsystems. To better secure operation of technological infrastructure, it is essential to build a comprehensive understanding of the atmospheric drivers, especially during solar active periods. Instrument measurements and climate models serve as valuable tools in obtaining information regarding the occurrence of space weather events; nonetheless, both sources exhibit quantitative and qualitative limitations. Data assimilation, an evolving technique, integrates measurements and model information to optimize the state estimations. This dissertation presents developments in a data assimilation algorithm known as Estimating Model Parameters from Ionospheric Reverse Engineering (EMPIRE), and its applications in investigating the atmospheric behaviors under varying solar conditions. EMPIRE is a data assimilation algorithm specifically designed for upper atmospheric driver estimation of neutral wind and ion drifts at user-defined spatial and temporal scales. The EMPIRE application in this work aims to contribute to a more comprehensive understanding of the effects of the NILE. EMPIRE utilizes the Kalman filter to optimize state calculations primarily based on electron density rates, provided by other data assimilation algorithms. Earlier runs of the algorithm used pre-defined values for the background state covariance cross time. To address model limitations under changing geomagnetic conditions, the algorithm is enhanced by concurrently updating the background state covariance during assimilation processes. Additionally, representation error is incor- porated as a component of the observation error, and error analysis is performed through a synthetic-data study. Previously, EMPIRE fused Fabry-Perot Interferometer (FPI) neutral wind measurements, demonstrating increased agreement with validation neutral wind data. In this work, this approach is extended to augment Coherent Scatter Radar (CSR) ion drift measurements from Super Dual Auroral Radar Network (SuperDARN), providing additional insights into EMPIRE’s estimated field-perpendicular ion motion. For an in-depth exploration of storm-related NILE, both EMPIRE and another data assimilation method, the Whole Atmosphere Community Climate Model with thermosphere and ionosphere eXtension coupled with Data Assimilation Research Testbed (WACCM-X + DART), is implemented for a storm event to test the proposed NILE driving mechanism. Furthermore, this dissertation introduces a Kalman smoother technique into the EMPIRE to enhance its ability to assess past storm events, and to explore the potential for algorithm improvements.
Show less

Title: Estimation of Platinum Oxide Degradation in Proton Exchange Membrane Fuel Cells
Creator: Ahmed, Niyaz Afnan
Date: 2024
Description: The performance and durability of Proton Exchange Membrane Fuel Cells (PEMFCs) can be significantly hampered due to the degradation of the...
Show moreThe performance and durability of Proton Exchange Membrane Fuel Cells (PEMFCs) can be significantly hampered due to the degradation of the platinum catalyst. The production of platinum oxide is a major cause of the degradation of the fuel cell system, negatively affecting its performance and durability. In order to predict and prevent this degradation, this research examines a novel method to estimate degradation due to platinum oxide formation and predict the level of platinum oxide coverage over time. Mechanisms of platinum oxide formation are outlined and two methods are compared for platinum oxide estimation. Linear regression and two Artificial Neural Network (ANN) models, including a Recurrent Neural Network (RNN) and Feed-forward Back Propagation Neural Network (FFBPNN), are compared for estimation. The estimation model takes into account the influence of cell temperature and relative humidity.Evaluation of relative errors (RE) and root mean square error (RMSE) illustrates the superior performance of RNN in contrast to GT-Suite and FFBPNN. However, both RNN and GT-Suite showcase an average error rate below 5% while the FFBPNN had a higher error rate of approximately 7%. The RMSE of RNN shows mostly less compared to FFBPNN and GT-Suite, however, at 50% training data, GT-Suite shows lowest RMSE. These findings indicate that GT-Suite can be a valuable tool for estimating platinum oxide in fuel cells with a relatively low RE, but the RNN model may be more suitable for real-time estimation of platinum oxide degradation in PEM fuel cells, due to its accurate predictions and shorter computational time. This comprehensive approach provides crucial insights for optimizing fuel cell efficiency and implementing effective maintenance strategies.
Show less

Title: Development of data assimilation for analysis of ion drifts during geomagnetic storms
Creator: Hu, Jiahui
Date: 2024
Description: The primary objective of this dissertation is to gain insight into geomagnetic storm effects at mid-latitudes induced by solar activity....
Show moreThe primary objective of this dissertation is to gain insight into geomagnetic storm effects at mid-latitudes induced by solar activity. Geomagnetic storms affect our everyday lives because they give rise to transient signal loss, data transmission errors, negatively impacting users of satellite navigation systems. The Nighttime Localized Ionospheric Enhancement (NILE) is a localized plasma enhancement that because it is not well understood, drives the design of satellite-based augmentationsystems. To better secure operation of technological infrastructure, it is essential to build a comprehensive understanding of the atmospheric drivers, especially during solar active periods. Instrument measurements and climate models serve as valuable tools in obtaining information regarding the occurrence of space weather events; nonetheless, both sources exhibit quantitative and qualitative limitations. Data assimilation, an evolving technique, integrates measurements and model information to optimize the state estimations. This dissertation presents developments in a data assimilation algorithm known as Estimating Model Parameters from Ionospheric Reverse Engineering (EMPIRE), and its applications in investigating the atmospheric behaviors under varying solar conditions. EMPIRE is a data assimilation algorithm specifically designed for upper atmospheric driver estimation of neutral wind and ion drifts at user-defined spatial and temporal scales. The EMPIRE application in this work aims to contribute to a more comprehensive understanding of the effects of the NILE. EMPIRE utilizes the Kalman filter to optimize state calculations primarily based on electron density rates, provided by other data assimilation algorithms. Earlier runs of the algorithm used pre-defined values for the background state covariance cross time. To address model limitations under changing geomagnetic conditions, the algorithm is enhanced by concurrently updating the background state covariance during assimilation processes. Additionally, representation error is incor- porated as a component of the observation error, and error analysis is performed through a synthetic-data study. Previously, EMPIRE fused Fabry-Perot Interferometer (FPI) neutral wind measurements, demonstrating increased agreement with validation neutral wind data. In this work, this approach is extended to augment Coherent Scatter Radar (CSR) ion drift measurements from Super Dual Auroral Radar Network (SuperDARN), providing additional insights into EMPIRE’s estimated field-perpendicular ion motion. For an in-depth exploration of storm-related NILE, both EMPIRE and another data assimilation method, the Whole Atmosphere Community Climate Model with thermosphere and ionosphere eXtension coupled with Data Assimilation Research Testbed (WACCM-X + DART), is implemented for a storm event to test the proposed NILE driving mechanism. Furthermore, this dissertation introduces a Kalman smoother technique into the EMPIRE to enhance its ability to assess past storm events, and to explore the potential for algorithm improvements.
Show less

repository.iit

Search the repository

Pages

Pages

Enabled Filters

Refine Results

Type

Date

Department

Subject

Creator

Part of

Rights