Digital Commons Network^™

Privacy Protection In Machine Learning Via Exploitation Of Constrained Adversarial Evasion, Brian Testa

Dissertations - ALL

Deep neural networks are extensively applied to real-world tasks. In many cases, the data feeding these tasks are human generated content, which emphasizes the criticality of privacy and data protection. The diverse modalities of this user data provide fertile ground for 3rd parties to monetize a user’s data. This work considers two such modalities: user speech when interacting with smart speaker voice assistants (VAs) and images shared with online service providers. In these cases, the user would like to support some form of machine learning (ML) inference without allowing others. A user interacting with a smart speaker would like the …

Cognitive Load Classification From Functional Near-Infrared Spectroscopy Data, Jiyang Wang

Dissertations - ALL

Predicting cognitive workload using physiological sensors, such as functional near-infrared spectroscopy (fNIRS), has gained significant attention in recent years. However, many existing methods are limited by small datasets and narrow brain channel coverage, restricting their generalizability across participants, tasks, and sessions. In this thesis, we introduce two innovative approaches aimed at overcoming these limitations: a CNN-BiGRU model with self-supervised label augmentation (SLA) and a block-wise domain adaptation method (BWise-DA). Our CNN-BiGRU-SLA model, grounded in cognitive load theory, leverages attention mechanisms and multi-label classification to predict working memory load (WML) and visual processing load (VPL) across multiple participants and sessions. Using …

Estranged Bodies, Race, And Sectarianism: An Historical Bioarchaeology Of Ottoman And French Beirut, Tony John Chamoun

Dissertations - ALL

Modern imperial formations involve a violent reconfiguring of difference that is inseparable from the everyday. Modern-world archaeology as a project that studies interlocking vicissitudes of modernity globally has done much to elucidate the varied social orders and politics of knowledge through which such reshuffling occurs. Yet largely missing from this conversation is the bioarchaeology of the Middle East and North Africa region; as such, the very bodies who shaped and were shaped by modern conditions in a major world circuit are also missing. To advance a more global and grounded investigation of modernity, I turn to the port city of …

Thermodynamics And Shear Flow Dynamics Of Block Copolymer Morphologies In Solution, Senyuan Liu

Dissertations - ALL

Block copolymers (BCPs) have been a focal point of research for several decades. They have been engineered to produce various morphologies by altering the combinations of chemically distinct polymer segments that interact selectively with their environment, such as in aqueous or organic solvents, or within a polymer matrix. The ability to tailor morphological features by manipulating polymer chain length, composition, and monomer chemistry has led to applications in advanced material manufacturing, catalysis, emulsification, environmental remediation, targeted drug delivery, gene therapy, and medical diagnostics. Consequently, understanding the relationship between copolymer architecture and the self-assembled morphologies at equilibrium, the pathways of structure …

Protocol For Monitoring Clonal Hematopoiesis In Transgenic Mouse Models Using Multispectral Imaging, Priyanka Khanna, Lauren B Ostermann, Shayaun Khazaei, Ran Zhao, Michael Andreeff, Rasoul Pourebrahim

Faculty, Staff and Student Publications

Clonal hematopoiesis involves the clonal expansion of hematopoietic cells, potentially progressing into hematological malignancies. Here, we present a protocol for the development and characterization of two mouse models designed to simulate clonal hematopoiesis and acute myeloid leukemia. We describe steps for model generation, monitoring clonal expansion, harvesting, fixation, and staining of bone marrow and spleen tissues. We specifically focus on the visualization and analysis of p53 mutant clonal expansions, providing a comprehensive protocol for studying these phenomena in mouse models. For complete details on the use and execution of this protocol, please refer to Pourebrahim et al.

Harmonic Impedance Modeling And Oscillation Analysis Of Modular Multilevel Converter, Yuhong Wang, Wensheng Chen, Shilin Gao, Jianquan Liao, Yangfan Cheng

Journal of System Simulation

Abstract: To facilitate rapid analysis of the oscillation stability mechanism in modular multilevel converter-based high voltage direct current (MMC-HVDC) systems and streamline the simulation process for determining MMC impedance characteristics, a simplified mathematical simulation model for MMC closed-loop impedance is developed using the harmonic state space method. This model considers various control strategies and includes both AC-side and DC-side impedance models. By applying a Nyquist criterion-based impedance analysis method, the stability mechanisms on the AC and DC sides of the MMC are examined. In addition, a data-driven oscillation stability analysis method is also proposed, leveraging a global sensitivity algorithm based …

Fail-Safe Logic Design Strategies Within Modern Fpga Architectures, Priya A. Bhakta

Electrical and Computer Engineering ETDs

Field Programmable Gate Arrays (FPGAs) are vulnerable to radiation-induced single event upsets (SEUs) and fault injection attacks, requiring the use of redundancy techniques such as fail-safe computing. Fail-safe computing refers to computing systems that revert to a non-operational safe state when a fault occurs. This work investigates circuit-level techniques and implements fail-safe computing processes as mitigation for SEUs and fault injection attacks on FPGAs. The analysis reveals vulnerabilities that exist in FPGAs over those in application-specific integrated circuits (ASIC); thus, requiring a more elaborate network of redundant circuits and checking logic. The reconfiguration capability of FPGAs adds complexity to fail-safe …

Pushyk V. Shiyanoff

All Decisions

In this no-grounds holdover proceeding, the landlord sought to evict the tenant. The tenant asserted that the apartment was subject to rent stabilization because the building had been previously altered to contain six apartments. The court agreed with the tenant. The court found that even though the landlord claimed one of the units had been removed, evidence suggested it had previously been rented as a residential unit, triggering rent stabilization for the entire building.

The Plasmodium Falciparum Ncr1 Transporter Is An Antimalarial Target That Exports Cholesterol From The Parasite's Plasma Membrane, Zhemin Zhang, Eva S Istvan, Daniel E Goldberg, Et Al.

2020-Current year OA Pubs

Malaria, a devastating parasitic infection, is the leading cause of death in many developing countries. Unfortunately, the most deadliest causative agent of malaria,

Developing Naturalistic Precision Functional Mapping For High-Density Diffuse Optical Tomography, Aahana Bajracharya

McKelvey School of Engineering Graduate Student Theses & Dissertations

Advancements in non-invasive imaging have enabled extensive research into higher-order cognitive functions of the brain. Functional Magnetic Resonance Imaging (fMRI) has provided valuable insights into brain networks underlying various cognitive processes. However, limitations such as confined environments and scanner noise pose challenges for fMRI to be broadly applicable across all demographics and experimental conditions, especially in speech and language studies. Because of this, there is a need for complementary techniques that can help understand the brain as it functions during everyday scenarios. High Density Diffuse Optical Tomography (HD-DOT) is an emerging functional near-infrared spectroscopy-based technique with comparative advantages over fMRI, …

Defining The Chondrocyte “Mechanome”: How Mechanosensitive Ion Channels Trpv4 And Piezo1 Drive Changes In Gene Expression To Affect Cell Behavior And Cartilage Health, Daniel Raymond Palmer

McKelvey School of Engineering Graduate Student Theses & Dissertations

Chondrocytes, the only native cell population of cartilage, utilize mechanosensitive transmembrane ion channels to sense mechanical forces. Transient Receptor Potential Vanilloid 4 (TRPV4) is activated by physiologic levels of load (resulting in ~10% tissue deformation) while PIEZO1 is activated by supraphysiologic levels of load (resulting in >50% tissue deformation). TRPV4 activation promotes the expression of a limited number of chondrogenic genes and the production of vital cartilaginous matrix proteins, while PIEZO1 activation drives cell death and senescence, cytoskeletal weakening, and production of catabolic mediators, which contribute to the development of osteoarthritis (OA). However, the path connecting activation of these channels …

Multimodal Representation Learning Frameworks For Modeling Progression And Heterogeneity In Alzheimer’S Disease, Sayantan Kumar

McKelvey School of Engineering Graduate Student Theses & Dissertations

Alzheimer’s Disease (AD) is the leading cause of dementia, characterised by cognitive and functional impairments that disrupt daily activities. Different clinical modalities such as neuroimaging biomarkers, cognitive assessments, fluid biomarkers and genetic data provide unique and complementary information, contributing to a more comprehensive understanding of disease progression and heterogeneity in disease characteristics. With recent advancements in computational capabilities, particularly in deep learning, multimodal representation learning frameworks aim to integrate diverse clinical modalities into a cohesive framework, capturing the most significant patterns within each modality. Existing data-driven multimodal representation learning frameworks in AD research have two major limitations. First, AD progresses …

Pan-Cancer Analysis Of Rna Dysregulation, Somatic Mutations, And Matrix Stiffness Using Bioinformatics Approaches, Gongyu Tang

McKelvey School of Engineering Graduate Student Theses & Dissertations

Cancer is a complex disease driven by genetic mutations, epigenetic modifications, and alterations in the tumor microenvironment. Understanding these intricate molecular mechanisms is crucial for advancing cancer diagnosis, prognosis, and treatment. RNA sequencing (RNA-seq) has emerged as a powerful tool to study cancer by providing comprehensive insights into gene expression, alternative splicing, and mutations at the transcriptome level. This dissertation leverages RNA-seq data to explore different dimensions of cancer biology through multi-omics integration and computational approaches. The research is divided into three main projects. First, I developed OncoDB, an interactive database to analyze gene expression, methylation patterns, and viral interactions …

The Encoding Of Decision Variables In Orbitofrontal Cortex (Ofc) Is Longitudinally Stable And Varies By Cortical Layer, Manning Zhang

McKelvey School of Engineering Graduate Student Theses & Dissertations

Economic choices entail computing the values of different options and then making choice by comparing values. Work in primates and rodents has shown that these mental processes rely on the OFC. Neurophysiology studies identified in OFC distinct groups of neurons encoding the value of individual options, the binary choice outcome, and the chosen value. These variables capture both the input and the output of the decision process, suggesting a decision circuit existing in OFC. A primary challenge for the field is to assess the structure and mechanisms governing this circuit. A first fundamental question is whether the circuit is stable …

Integrating Physics And Deep Learning For Transmission-Less Attenuation Compensation In Clinical Spect, Zitong Yu

McKelvey School of Engineering Graduate Student Theses & Dissertations

Attenuation compensation (AC) is beneficial for visual interpretation and serves as a prerequisite for quantification tasks in single-photon emission computed tomography (SPECT) imaging. However, conventional AC methods typically rely on a separate X-ray CT component, leading to multiple challenges. In this context, studies have shown that scatter-window projection data contains information to estimate the attenuation distribution. Furthermore, large amounts of SPECT emission data in both photopeak and scatter-energy windows and corresponding CT scans are available. The ability of deep learning (DL) to model complex relationships by leveraging large datasets motivates its use for transmission-less AC in SPECT. Building upon the …

Developing Naturalistic Precision Functional Mapping For High-Density Diffuse Optical Tomography, Aahana Bajracharya

McKelvey School of Engineering Graduate Student Theses & Dissertations

Advancements in non-invasive imaging have enabled extensive research into higher-order cognitive functions of the brain. Functional Magnetic Resonance Imaging (fMRI) has provided valuable insights into brain networks underlying various cognitive processes. However, limitations such as confined environments and scanner noise pose challenges for fMRI to be broadly applicable across all demographics and experimental conditions, especially in speech and language studies. Because of this, there is a need for complementary techniques that can help understand the brain as it functions during everyday scenarios. High Density Diffuse Optical Tomography (HD-DOT) is an emerging functional near-infrared spectroscopy-based technique with comparative advantages over fMRI, …

Defining The Chondrocyte “Mechanome”: How Mechanosensitive Ion Channels Trpv4 And Piezo1 Drive Changes In Gene Expression To Affect Cell Behavior And Cartilage Health, Daniel Raymond Palmer

McKelvey School of Engineering Graduate Student Theses & Dissertations

Chondrocytes, the only native cell population of cartilage, utilize mechanosensitive transmembrane ion channels to sense mechanical forces. Transient Receptor Potential Vanilloid 4 (TRPV4) is activated by physiologic levels of load (resulting in ~10% tissue deformation) while PIEZO1 is activated by supraphysiologic levels of load (resulting in >50% tissue deformation). TRPV4 activation promotes the expression of a limited number of chondrogenic genes and the production of vital cartilaginous matrix proteins, while PIEZO1 activation drives cell death and senescence, cytoskeletal weakening, and production of catabolic mediators, which contribute to the development of osteoarthritis (OA). However, the path connecting activation of these channels …

Reimagining The Machine Vision Pipeline In Cyber Physical Systems For Trustworthiness And Efficiency, Adith Boloor

McKelvey School of Engineering Graduate Student Theses & Dissertations

Cyber-physical systems (CPS), including autonomous vehicles, drones, and mobile robots, rely on intricate sensors, actuators, and machine learning algorithms to perceive the physical world and execute actions within their surroundings. In the context of vision-driven CPS, achieving this demands processing a substantial volume of visual data captured by on-board cameras. The data is subsequently channeled through digital processors and harnessed by deep neural networks for tasks such as image classification, object detection, and depth perception. This data-centric, machine-vision-infused CPS fosters intelligent decision-making, thereby enhancing overall system performance. Trustworthiness, encompassing the robustness of the entire machine-vision pipeline, and system-level efficiency are …

Multimodal Representation Learning Frameworks For Modeling Progression And Heterogeneity In Alzheimer’S Disease, Sayantan Kumar

McKelvey School of Engineering Graduate Student Theses & Dissertations

Alzheimer’s Disease (AD) is the leading cause of dementia, characterised by cognitive and functional impairments that disrupt daily activities. Different clinical modalities such as neuroimaging biomarkers, cognitive assessments, fluid biomarkers and genetic data provide unique and complementary information, contributing to a more comprehensive understanding of disease progression and heterogeneity in disease characteristics. With recent advancements in computational capabilities, particularly in deep learning, multimodal representation learning frameworks aim to integrate diverse clinical modalities into a cohesive framework, capturing the most significant patterns within each modality. Existing data-driven multimodal representation learning frameworks in AD research have two major limitations. First, AD progresses …

Integrating Physics And Deep Learning For Transmission-Less Attenuation Compensation In Clinical Spect, Zitong Yu

McKelvey School of Engineering Graduate Student Theses & Dissertations

Attenuation compensation (AC) is beneficial for visual interpretation and serves as a prerequisite for quantification tasks in single-photon emission computed tomography (SPECT) imaging. However, conventional AC methods typically rely on a separate X-ray CT component, leading to multiple challenges. In this context, studies have shown that scatter-window projection data contains information to estimate the attenuation distribution. Furthermore, large amounts of SPECT emission data in both photopeak and scatter-energy windows and corresponding CT scans are available. The ability of deep learning (DL) to model complex relationships by leveraging large datasets motivates its use for transmission-less AC in SPECT. Building upon the …

The Encoding Of Decision Variables In Orbitofrontal Cortex (Ofc) Is Longitudinally Stable And Varies By Cortical Layer, Manning Zhang

McKelvey School of Engineering Graduate Student Theses & Dissertations

Economic choices entail computing the values of different options and then making choice by comparing values. Work in primates and rodents has shown that these mental processes rely on the OFC. Neurophysiology studies identified in OFC distinct groups of neurons encoding the value of individual options, the binary choice outcome, and the chosen value. These variables capture both the input and the output of the decision process, suggesting a decision circuit existing in OFC. A primary challenge for the field is to assess the structure and mechanisms governing this circuit. A first fundamental question is whether the circuit is stable …

Reimagining The Machine Vision Pipeline In Cyber Physical Systems For Trustworthiness And Efficiency, Adith Boloor

McKelvey School of Engineering Graduate Student Theses & Dissertations

Cyber-physical systems (CPS), including autonomous vehicles, drones, and mobile robots, rely on intricate sensors, actuators, and machine learning algorithms to perceive the physical world and execute actions within their surroundings. In the context of vision-driven CPS, achieving this demands processing a substantial volume of visual data captured by on-board cameras. The data is subsequently channeled through digital processors and harnessed by deep neural networks for tasks such as image classification, object detection, and depth perception. This data-centric, machine-vision-infused CPS fosters intelligent decision-making, thereby enhancing overall system performance. Trustworthiness, encompassing the robustness of the entire machine-vision pipeline, and system-level efficiency are …

Spatial Insight Of Kidney Sex-Dimorphism And Tumor Evolution, Chia-Kuei Mo

McKelvey School of Engineering Graduate Student Theses & Dissertations

The landscape of spatial biology research has transformed drastically in recent years allowing a comprehensive understanding of cellular organization and molecular mechanisms in various biological context. This thesis presents three complementary studies that advance our understanding of spatial organization in normal tissue biology and tumor evolution, utilizing cutting-edge spatial multi-omics approaches. The first study presents a comprehensive investigation of kidney sexual dimorphism across the lifespan, integrating six distinct technological platforms: single-nucleus transcriptomics, epigenomics, Visium spatial transcriptomics, Xenium single-cell resolution spatial transcriptomics, conventional immunofluorescence imaging, and co-detection by indexing (CODEX) multiplexed protein imaging. This multi-modal approach reveals that proximal tubules exhibit …

Discrete Element Method For Mining Industrial Applications: Solving The Clumping Problem Caused By Bulk Cohesive Particles In An Iron Ore Beneficiation Process, Shuang Liu

McKelvey School of Engineering Graduate Student Theses & Dissertations

The discrete element method (DEM) is a widely used numerical approach for simulating and analyzing the behavior of granular materials in various applications. This research focuses on the iron ore beneficiation process, addressing the clumping issue in cohesive iron ores from existing crushing processes through a novel double-layered disperser design. The significance of this study lies in its potential to enhance the efficiency of the existing crushing phase and improve the linkage to subsequent purification stages that target smaller-sized iron ore pellets, thereby enhancing the overall efficiency of the ore beneficiation process and achieving significant energy savings. This research adopts …

Pan-Cancer Analysis Of Rna Dysregulation, Somatic Mutations, And Matrix Stiffness Using Bioinformatics Approaches, Gongyu Tang

McKelvey School of Engineering Graduate Student Theses & Dissertations

Cancer is a complex disease driven by genetic mutations, epigenetic modifications, and alterations in the tumor microenvironment. Understanding these intricate molecular mechanisms is crucial for advancing cancer diagnosis, prognosis, and treatment. RNA sequencing (RNA-seq) has emerged as a powerful tool to study cancer by providing comprehensive insights into gene expression, alternative splicing, and mutations at the transcriptome level. This dissertation leverages RNA-seq data to explore different dimensions of cancer biology through multi-omics integration and computational approaches. The research is divided into three main projects. First, I developed OncoDB, an interactive database to analyze gene expression, methylation patterns, and viral interactions …

Discrete Element Method For Mining Industrial Applications: Solving The Clumping Problem Caused By Bulk Cohesive Particles In An Iron Ore Beneficiation Process, Shuang Liu

McKelvey School of Engineering Graduate Student Theses & Dissertations

The discrete element method (DEM) is a widely used numerical approach for simulating and analyzing the behavior of granular materials in various applications. This research focuses on the iron ore beneficiation process, addressing the clumping issue in cohesive iron ores from existing crushing processes through a novel double-layered disperser design. The significance of this study lies in its potential to enhance the efficiency of the existing crushing phase and improve the linkage to subsequent purification stages that target smaller-sized iron ore pellets, thereby enhancing the overall efficiency of the ore beneficiation process and achieving significant energy savings. This research adopts …

Retrospective Derivation Of A Causal Pathway For Diabetic Ketoacidosis In Adult Patients With Type 2 Diabetes Mellitus, Jeffrey A Kline, Nicholas A Wesner, Amina T Sharif, Richard T Griffey, Phillip D Levy, Robert D Welch, George Grunberger

2020-Current year OA Pubs

BACKGROUND: Type 2 ketone-prone diabetes mellitus (T2KPDM) is thought to occur in men of African descent, with obesity who experienced prolonged hyperglycemia; the role of medication non-adherence as a contributing cause remains unstudied.

RESEARCH DESIGN AND METHODS: This was a retrospective study of unique adults (>18 years) who sought emergency care one of four hospitals in the greater Detroit area. Patients were identified on the basis of a laboratory order for a ß-hydroxybutyrate concentration. Two research coordinators abstracted 119 data fields. Patients were divided into four phenotypes: (1) no prior DM, (2) type 2 DM without prior ketosis, (3) …

Pancreaticoduodenectomy In Trauma Patients With Grade Iv-V Duodenal Or Pancreatic Injuries: A Post Hoc Analysis Of An East Multicenter Trial, Rachel Leah Choron, Rami Al-Aref, Jordan Michael Kirsch, Et Al.

2020-Current year OA Pubs

INTRODUCTION: The utility of pancreaticoduodenectomy (PD) for high-grade traumatic injuries remains unclear and data surrounding its use are limited. We hypothesized that PD does not result in improved outcomes when compared with non-PD surgical management of grade IV-V pancreaticoduodenal injuries.

METHODS: This is a retrospective, multicenter analysis from 35 level 1 trauma centers from January 2010 to December 2020. Included patients were ≥15 years of age with the American Association for the Surgery of Trauma grade IV-V duodenal and/or pancreatic injuries. The study compared operative repair strategy: PD versus non-PD.

RESULTS: The sample (n=95) was young (26 years), male (82%), …

Re: Butte Mine Flooding Operable Unit Berkeley Pit And Discharge Pilot Project – November 2024 Monthly Discharge Monitoring Report, Mark Thompson, Dave Griffis

Silver Bow Creek/Butte Area Superfund Site

No abstract provided.

Testing The Bullet Dwarf Collision Scenario In The Ngc 1052 Group Through Morphologies And Stellar Populations, Yimeng Tang, Aaron J. Romanowsky, Pieter G. Van Dokkum, T. H. Jarrett, Kevin A. Bundy, Maria Luisa Buzzo, Shany Danieli, Jonah S. Gannon, Michael A. Keim, Seppo Laine, Zili Shen

Faculty Research, Scholarly, and Creative Activity

NGC 1052-DF2 and -DF4 are two ultradiffuse galaxies that have been reported as deficient in dark matter and associated with the same galaxy group. Recent findings suggest that DF2 and DF4 are part of a large linear substructure of dwarf galaxies that could have been formed from a high-velocity head-on encounter of two gas-rich galaxies, known as a “bullet dwarf” collision. Based on new observations from the Hubble Space Telescope, combined with existing imaging from the u band to mid-infrared, we test the bullet dwarf scenario by studying the morphologies and stellar populations of the trail dwarfs. We find no …