Digital Commons Network^™

Improving Cancer Diagnosis And Patient Outcomes With Deep Learning Models, Mariana Arriz-Jorquiera

USF Tampa Graduate Theses and Dissertations

Cancer care depends on timely and reliable decisions, from detection and diagnosis to treatment planning and patient monitoring. These decisions are often made under uncertainty because medical images and healthcare data may be noisy, incomplete, or difficult to interpret. In breast cancer imaging, ultrasound is widely used because it is safe, accessible, and complementary to other imaging modalities. However, variations in image quality, acquisition conditions, and noise can obscure lesion boundaries and texture, affecting human interpretation and artificial intelligence reliability. This dissertation develops deep learning, image-analysis, and optimization methods to improve healthcare decisions under imperfect information. Its primary focus is …

From Shortage To Strategy: Optimizing Scarce Resource Allocation In Healthcare, Daniela Duran Cantarino

USF Tampa Graduate Theses and Dissertations

The allocation of scarce medical resources presents one of the most consequential challenges in health care, requiring decisions that balance equity and efficiency under pervasive uncertainty. This dissertation develops optimization-based frameworks for two critical domains, organ transplantation and emergency medical services, and demonstrates how operations research methods can improve upon current practice along both dimensions simultaneously.

In the domain of deceased-donor kidney allocation, two complementary frameworks are proposed. The first is a bi-objective stochastic optimization model that jointly maximizes post-transplant survival (efficiency) and prioritizes candidates with the highest pre-transplant mortality risk (equity), subject to regional chance constraints on graft failure …

Toward Intelligent Machines: Conscious Learning And Hardware-Accelerated Ai, Pavia Bera

USF Tampa Graduate Theses and Dissertations

Artificial intelligence systems excel at narrow, well-defined tasks but remain brittle at theboundaries of their training distributions: they cannot quantify uncertainty, adapt continuously to non-stationary data, or operate efficiently on energy-constrained hardware. This dissertation addresses these limitations through four coordinated contributions spanning probabilistic learning algorithms, biologically inspired temporal memory, cross-entity warning propagation via distributed associative memory, and spintronic processing-in-memory.

The first contribution introduces Boosted Bayesian Neural Networks (BBNNs), which extend standard mean-field variational inference by iteratively constructing a mixture posterior through Boosting Variational Inference. On five clinical medical classification datasets, BBNNs achieve superior uncertainty calibration—lower Negative Log-Likelihood and Expected Calibration …

Autoencoders As Classifiers Trained On Single Sources Of Radiographic Images For Generalizability Across Unseen Sources, Ryan M. Putney

USF Tampa Graduate Theses and Dissertations

Artificial neural networks trained to classify X-ray images according to disease state will learn to distinguish between technical or procedural variations in the images rather than features relevant to the disease. For example, the model might learn to recognize the machine that captured the image or to distinguish an image taken while the patient is lying supine or standing upright. The present work is aimed at using autoencoders as a generalizable classifier to detect COVID vs pneumonia (PNA) X-ray images. The first experiment was to find an architecture for a fully convolutional autoencoder (CAE) and for a convolutional autoencoder with …

Multi-Modal Survival Prediction On Breast Cancer Mammograms And Enabling Multi-Institutional Collaborations With Federated Learning, Nikolas Koutsoubis

USF Tampa Graduate Theses and Dissertations

Cancer remains one of the leading causes of mortality worldwide, and machine learning is well positioned to leverage the large volumes of routinely collected clinical and imaging data to improve patient outcomes. Realizing this potential at scale requires three capabilities that current practice does not yet deliver in combination: models that integrate the multiple data modalities clinicians use, training procedures that respect institutional data-sharing constraints, and tooling that enables sharing of de-identified data when federated approaches are not sufficient. This dissertation develops methods and infrastructure addressing each of these capabilities in the context of breast cancer and oncology more broadly. …

Efficient, Secure Learning And Resource Management For Wireless Networks, Jiahao Xue

USF Tampa Graduate Theses and Dissertations

The rapid advancement of wireless networks has enabled a wide range of applications such as federated learning (FL) and Internet of Things, with supporting technologies like radio frequency identification (RFID). However, resource management and security challenges in wireless networks remain insufficiently addressed for these representative applications. In this dissertation, we first investigate wireless resource management for FL, then study novel attack and defense mechanisms for RFID systems. Finally, we analyze the performance and security of artificial intelligence (AI)-enabled next generation wireless networks.

First, we investigate the client assignment and bandwidth allocation problem for FL in a multi-provider wireless setting. Unlike …

Building Trustworthy Information Systems: A Unified Framework For Comparative Risk Detection, Parisa Momeni

USF Tampa Graduate Theses and Dissertations

Risk detection in large scale information systems increasingly depends on heterogeneous data generatedby both centralized and distributed ecosystems. While centralized systems provide curated and validated reports, distributed environments produce large-scale and real-time observational evidence. Existing computational approaches analyze these ecosystems in isolation, limiting systematic comparison of risk repre-sentations across heterogeneous sources.

This dissertation presents a unified computational framework for comparative risk detection across centralized and distributed information systems. The framework provides a domain independent methodology for transforming heterogeneous risk reporting data into comparable multidimensional representations. To enable interpretable comparison of heterogeneous risk distributions, this work introduces the Geometric Overlap Score …

Seeing Around Corners: An Indirect Light Transport Decomposition Framework For Fusing Physics And Learned Priors, Fadlullah Raji

USF Tampa Graduate Theses and Dissertations

Non-line-of-sight (NLOS) imaging, or seeing around corners, is the ability to recover information about objects hidden from direct view and remains one of the most compelling challenges in computational imaging. Existing approaches fall broadly into two categories: active and passive. Active time-resolved methods achieve impressive performance but require specialized pulsed laser hardware and ultrafast detectors, making them expensive and limited by slow measurement acquisition. Passive NLOS methods, which instead exploit the subtle soft shadows (penumbrae) cast by a hidden scene onto a visible matte surface, have recently emerged as a practical alternative. However, existing passive approaches are largely limited to …

Language Models For Oncology Clinical Text: How Model Architecture And Data Strategies Shape Tumor Phenotype Extraction And Disease Progression Detection, Thanh Duong

USF Tampa Graduate Theses and Dissertations

The rapid growth of electronic health records (EHRs) has created new opportunities to apply machine learning to clinical data.However, a large portion of important clinical information is still stored in unstructured text, such as pathology reports, radiology reports, and longitudinal clinical notes.These documents contain key details about tumor characteristics, diagnoses, treatments, and patient outcomes.Extracting structured and useful information from this text is challenging due to complex medical language, varied document formats, and the need to combine information across multiple reports over time.This dissertation studies how language models can be designed and adapted to better extract and use oncology-specific information from …

Mapping Seafloor Habitats And Characterizing Demersal Reef Fish Communities On The West Florida Shelf Across Multiple Spatial Scales Using Multibeam Sonar And Underwater Video, Alexander R. Ilich

USF Tampa Graduate Theses and Dissertations

In this dissertation, I examine the role of spatial scale in deriving seafloor terrain attributes, modeling benthic habitat, and analyzing ecological patterns of the demersal reef fish communities on the West Florida Shelf (WFS), with implications for improving predictive habitat mapping and understanding fish-environment relationships. Spatial scale plays a central role in benthic habitat mapping and ecological inference, influencing how terrain attributes are derived, how habitats are modeled, and how ecological responses are manifested. This work integrates multiscale terrain analysis, acoustic remote sensing, and community ecology to evaluate how spatial scale influences both physical habitat representation and ecological analysis.

Chapter …

Portable Trace Gas Analysis With Multipass Cavity Raman Scattering, Charuka Muktha Arachchige

USF Tampa Graduate Theses and Dissertations

The continual increase in the production and utilization of chemicals across a growing range of industrial, environmental, and medical applications necessitates accurate and efficient measurement of molecular composition. Trace gas detection is particularly critical in industrialized settings, where the identification of colorless and odorless gases is essential for ensuring operational safety and regulatory compliance. Recent advances in optical instrumentation have positioned spontaneous Raman scattering as a promising analytical technique for trace gas analysis, with multipass cavity enhancement providing a cost-effective pathway to high detection sensitivity. However, existing implementations remain confined to laboratory-scale optical table setups, limiting their applicability in field …

Designing For Fitness And Resilience In Human Artificial Intelligence Systems, Thomas R. Gill

USF Tampa Graduate Theses and Dissertations

Over the last decade, new artificial intelligence models, frameworks, and applications have emerged that have radically changed the socio-technical landscape. These new technologies represent both great opportunities and great potential risks. Due to the recency of their emergence, there is little known about the longevity and long-term repercussions of artifacts powered by new artificial intelligence technology. In this dissertation, I evaluate A.I. and A.I.-based artifacts through the lens of the fitness-utility model, a tool developed by Gill and Hevner (2013) for the evaluation of an artifact's ability to survive and reproduce over long time horizons. In so doing, I also …

Preparing Preservice Teachers: Developing Skills For Working With Students With Disabilities In The Social Studies Classroom, Melissa Waugh

USF Tampa Graduate Theses and Dissertations

This study was aimed towards investigating the extant research, best practices, and teacher preparation standards related to working with students with disabilities in the social studies classroom. The study was guided using three research questions: Are students with disabilities’ learning needs reflected in the National Council for Social Studies National Standards for the Preparation of Social Studies Teachers?; How do preservice social studies teacher programs prepare their candidates for teaching students with disabilities as evidenced by Social Science Education methods course syllabi?; and, What might a college-level course aimed towards social studies teaching methods centering around students with disabilities’ learning …

Photoluminescent Studies Of Metal-Organic Frameworks For Small Molecule Sensing, Julia T. Dematteo

USF Tampa Graduate Theses and Dissertations

Metal–organic frameworks (MOFs) are highly tunable nanoporous materials with broad applications in gas separation, drug delivery, light harvesting, and chemical sensing. The incorporation of photoactive components into MOFs enables the development of optical sensors capable of detecting a wide range of analytes. Porphyrins and metalloporphyrins are particularly attractive for such applications due to their high molar absorptivity, strong emission quantum yields, and versatile interactions with chemical species.

In this work, two porphyrinic MOFs, PCN-222 and MOF1(Cd), are investigated. Both frameworks utilize tetrakis(4-carboxyphenyl)porphyrin (TCPP) as the organic linker, imparting optical functionality to the materials. Integration of TCPP into the framework induces …

Risk-Aware Deep Reinforcement Learning In Portfolio Optimization, Shimin Zhang

USF Tampa Graduate Theses and Dissertations

Portfolio optimization is important in investment, with the purpose of maximizing profit whilemanaging risk. The traditional portfolio optimization method, Modern Portfolio Theory, is based on the assumption that returns are normally distributed, which is always violated in practical financial market settings. In addition, traditional approaches to portfolio optimization are mostly designed as static optimizations and fail to account for the nonlinearity and non-stationarity of contemporary financial markets. Recent developments in deep reinforcement learning have attracted considerable attention for treating portfolio optimization as a sequential decision-making process. Traditional portfolio optimization methods usually evaluate risk based on the variance of the returns. …

Van Der Waals Interactions Through The Lens Of 2d Materials With Emphasis On Many-Body Effects, Thi Xuan Diem Dang

USF Tampa Graduate Theses and Dissertations

This dissertation addresses the long-standing challenge of accurately and efficiently modeling long-range electron correlations, with a particular focus on van der Waals (vdW) interactions. These interactions play an essential role in predicting the structure, stability, and functionality of molecular systems and condensed matter materials. Although Density Functional Theory (DFT) has become a robust cornerstone of computational materials science, traditional formulations of the theory frequently fail to capture dispersive interactions, leading to significant errors in determining interlayer distances, binding energies, and stacking configurations. These limitations become particularly pronounced in layered materials and vdW heterostructures (HSTs), where weak long-range correlations are sensitive …

A Portable Device To Elicit Sensory Trick And Monitor Forearm Muscle Activity In Pianists With Musician’S Dystonia, Josef Franczak

USF Tampa Graduate Theses and Dissertations

Musician’s Dystonia (MD) is a condition that can be acquired in professional musicians,causing involuntary contractions and the loss of fine motor control. The most common pharmaceutical treatment, botulinum toxin injections, has been shown to have side effects of muscle weakness and atrophy. The sensory trick, shown to temporarily relieve symptoms of dystonia, has potential to be used as an alternative or alongside current therapies, allowing patients with MD to return to playing music professionally. To systematically deploy the sensory trick and monitor forearm muscle activity, the first prototype of the Wearable Sensorimotor Integrator (wSMI) has been developed.

The wSMI is …

Photophysical Behavior Of Ruthenium-Based Complexes: A Study Of Ligand Variation And Encapsulation In Metal-Organic Frameworks, Delaney S. Sellers

USF Tampa Graduate Theses and Dissertations

Ruthenium (II) polypyridyl complexes have emerged as a prominent class of compounds in inorganic photochemistry due to their structural robustness, reversible redox chemistry, and tunable excited-state dynamics. Central to the investigation of these systems is the ability to manipulate the triplet metal-to-ligand charge transfer state (3MLCT). In the context of photoactivated chemotherapy (PACT), the strategic deactivation of this MLCT state to populate a dissociative triplet ligand field state (3LF) can enable the controlled, light-triggered release of caged cytotoxic agents. This work focuses on understanding the caging mechanism of various ligands to optimize the eventual delivery of cytotoxic ligands. In this …

Development Of A Coupled Hydrological-Nitrogen Modeling Framework For Urban Stormwater Pond Analysis, Tione Grant

USF Tampa Graduate Theses and Dissertations

Urbanization impacts the hydrological and biogeochemical processes that regulate water quality in downstream ecosystems, leading to increased stormwater volume and nitrogen loads in receiving aquatic environments. Although stormwater ponds are designed as simple hydraulic storage basins, they function more like managed ecosystems within urban drainage systems, offering both hydraulic attenuation and passive nutrient treatment. However, nitrogen removal efficiency varies significantly and depends on interactions between hydrological and biogeochemical factors. This thesis details the development, calibration, validation, and sensitivity analysis of a coupled hydrological-nitrogen model framework for Aaran’s Pond, an urban stormwater pond in Hillsborough County, Florida.

A process-based water budget …

Trophic Structure Of Reef-Fish Assemblages In The Eastern Gulf Of Mexico Varies By Depth And Reef Type, Jessica L. Van Vaerenbergh

USF Tampa Graduate Theses and Dissertations

Globally, marine ecosystems are increasingly threatened by anthropogenic stressors, leading to reef degradation, changes in reef-fish assemblages, and altered trophic structure. In this study, I investigated reef-fish underwater visual survey data collected over a decade (2013-2023) from paired natural and artificial reefs in the eastern Gulf of Mexico. I applied the novel Normalized Reef Status Index (NRSI) to evaluate how biomass was distributed across the three trophic categories (i.e., Lower Trophic Position (LTP; 2.0-2.9), Central Trophic Position (CTP; 3.0-3.9), and Upper Trophic Position (UTP; ≥4.0). This index is a metric to assess the ecological status of reef systems by investigating …

Development And Evaluation Of The East Florida Coastal Ocean Model (Efcom), Orion B. Scharton-Witmer

USF Tampa Graduate Theses and Dissertations

Florida’s Atlantic coast has abundant resources in beaches, state parks, aquatic preserves, and wildlife refuges as well as densely populated coastal cities in low lying areas that are all subjected to natural disaster threats from the oceans. There is an urgent need for a high-resolution coastal ocean circulation model for this region. Based on the application of the Finite Volume Community Ocean Model (FVCOM), an East Florida Coastal Ocean Model (EFCOM) is developed, downscaling from the deep ocean, across the continental shelf, and into the estuaries. It employs an unstructured grid with horizontal resolution varying from 3 kilometers at the …

A System Dynamics Perspective Of User Adaptation: The Case Of Exoskeleton Technology, Enzo Novi Migliano

USF Tampa Graduate Theses and Dissertations

Technology implementation is a common occurrence in organizations to which workers must adapt. Although research on workers' technology acceptance is widely spread, research on the workers' technology adaptation process is less represented in the literature. The discrepancy in the literature permeates research on new technologies currently being implemented in organizations, such as occupational exoskeletons. As in other technologies, the variance research perspective dominates the literature on exoskeleton implementations. The current thesis aims to foster process research by formalizing the coping model of user adaptation (Beaudry & Pinsonneault, 2005) and exploring research questions regarding workers' adaptation to general technology and exoskeletons. …

Amending Planting Media With Typha-Derived Biochar For Sustainable Horticulture, Elissa Touma

USF Tampa Graduate Theses and Dissertations

Constructed wetlands are widely used to improve water quality by removing excess nutrients from wastewater and stormwater. Nutrients are stored within wetland vegetation, particularly in fast-growing plants such as Typha. Harvesting vegetation provides an opportunity to permanently remove these nutrients while generating biomass suitable for beneficial reuse. Converting harvested biomass into value-added products such as compost and biochar can support nutrient recovery and sustainable biomass management. In addition, increasing concerns regarding the environmental impacts of peat extraction have created interest in renewable alternatives for horticultural growing media. Therefore, the objective of this study was to evaluate the potential of harvested …

Exploring Culturally Responsive Leadership Practices Among Female Leaders In Title I Schools: Strategies For Enhancing Student Success, Tara K. Indar

USF Tampa Graduate Theses and Dissertations

The academic achievement gap between White students and diverse student populations remains a persistent and prominent disparity in schools across the United States. While this issue affects schools nationwide, it is particularly significant in Title I schools, where many students from diverse, lower-socioeconomic-status backgrounds are served. Although a substantial body of literature examines this gap and highlights the positive impact of culturally responsive leadership, there remains a limited understanding of which specific practices are most effective in improving students’ academic outcomes. This study explores the perceptions of female school leaders in Title I schools regarding the strategies they believe contribute …

Modeling And Mitigating Stale Hardcoded Secrets In Version Control History, Katarina Valentine Capalbo

USF Tampa Graduate Theses and Dissertations

Hardcoded secret vulnerabilities remain a growing and persistent problem that can be difficult to mitigate once exposed. While there has been significant research and advancements in identifying and preventing hardcoded secrets, less attention has been given to understanding the risks of leaving these secrets behind in version control history, as well as analyzing the sanitization of hardcoded secrets. These challenges present a growing gap between the risk and ability to mitigate and sanitize existing secret exposures. This thesis begins to address this gap by formalizing a threat model for hardcoded secrets that persist in Git version control history after the …

Morphodynamic Evolution Of Midnight Pass, West-Central Florida, Sydney C. Scott

USF Tampa Graduate Theses and Dissertations

This study takes advantage of a rare opportunity provided by the reopening of Midnight Pass due to two consecutive hurricanes in 2024, to examine the morphologic evolution of a newly reopened inlet and subsequent beach-inlet interactions. Historical aerial photos since the 1940s were collected and analyzed to understand the last closure of Midnight Pass, and to gain knowledge on the future of the newly reopened inlet. Time-series topography and bathymetry survey of the beach-inlet system were conducted bimonthly to quarterly for the first year and a half after the reopening, with the goals of quantifying the morphologic changes and understanding …

Quantitative Determination Of Nano Delivered Kras Sirna, Axel S. Deraspe

USF Tampa Graduate Theses and Dissertations

This study focuses on the development and validation of analytical methods for the detection of therapeutic KRAS-targeting siRNA in biological systems. Physicochemical characterization showed that p5RHH-formulated siRNA nanoparticles exhibit a controlled nanoscale size distribution, moderate polydispersity, and a negative zeta potential, suggesting colloidal stability and potential suitability for cellular uptake. Scanning electron microscopy confirmed a predominantly spherical morphology with a relatively uniform size distribution, consistent with dynamic light scattering measurements.

To enable accurate siRNA quantification, stem–loop reverse transcription (SL-RT) coupled with SYBR Green–based quantitative Polymerase Chain Reaction (qPCR) was optimized. Key parameters, including a primer concentration of 100 nM and …

Students In Motion: How Educators Enact Movement In Elementary Mathematics, Katherine E. Nash

USF Tampa Graduate Theses and Dissertations

This study used an exploratory sequential mixed-methods design to examine how elementary mathematics teachers report using physical activity in their instructional practice. The purpose of the study was to develop, provide evidence of validity, and implement a questionnaire to document the types of movement teachers use and when these practices occur in relation to mathematics instruction. In Phase 1, qualitative interviews were conducted with eight elementary mathematics teachers to identify common movement types and inform instrument development. The eight movement types that were identified are non-academic fine-motor, non-academic gross-motor, non-academic music-elicited, non-academic environmental, academic fine-motor, academic gross-motor, academic music-elicited, and …

Measures Of Linguistic Complexity And Vocabulary Diversity As Indicators Of Oral Language Production Growth Of 4th Grade Spanish-Speaking Els, Carla T. Zayas-Santiago

USF Tampa Graduate Theses and Dissertations

As fourth-grade students transition from learning to read to reading to learn, their ability to use language for academic purposes becomes increasingly important, particularly for Spanish-speaking English Learners (ELs). However, students often have limited opportunities to produce extended oral language during instruction. Students’ oral language production reflects both structural aspects of language use, captured through measures of linguistic complexity, and lexical aspects of language use, captured through measures of vocabulary diversity. Although prior studies have examined transcript-derived measures of oral language production, limited research has examined how these measures predict English proficiency, capture growth over time, and respond to different …

The Architecture Of Chemical Understanding: How Representation Selection, Ordering, And Feature Salience Shape Students' Conceptions In Chemistry, Isaiah Nelsen

USF Tampa Graduate Theses and Dissertations

Chemical representations play a central role in how students learn and reason aboutchemistry. While representations are routinely used in chemistry classrooms, students often struggle to extract relevant information, connect representations to underlying chemical principles, and apply this information to problem-solving. This dissertation investigates the role of representations in supporting students’ conceptual understanding and reasoning across general and organic chemistry through three qualitative studies.

The first study in this work explores how variation in chemical representation influences how students understand foundational chemistry principles. This study sought to characterize how second semester general chemistry students approach dipole–dipole interaction tasks with four distinct …