2025 UMKC Dissertations - Freely Available Online
Permanent URI for this collection
Browse
Recent Submissions
Item Illustrating the clinical landscape of Mucorales infection: a comprehensive examination of demographic characteristics, regional variation, length of stay, and readmission rates of cases in the United States(2025) Jones, Andrew Paul; Allsworth, Jenifer E.Mucormycosis is a rare but devastating fungal infection that primarily afflicts immunocompromised patients including those with hematological malignancy, solid organ and bone marrow transplants, and diabetes mellitus. The objective of this study was to assess the prevalence and clinical burden of mucormycosis among hospitalized patients in the United States. The three studies conducted analyzed the Oracle Health Facts® database, a deidentified electronic health record resource, which includes more than 750 participating healthcare facilities, 500 million unique patient encounters, 69 million patients, and 4.7 billion laboratory results between 2000 and 2018. All inpatient hospitalizations were examined for documentation of mucormycosis using an ICD-9-CM code of 117.7 or ICD-10-CM codes of B46.0-B46.9. In study 1, we estimated the prevalence of mucormycosis-related hospitalizations nationally, by census region and demographic characteristics, and described temporal trends. In study 2, we conducted a matched case-control study design to estimate the association of mucormycosis on length of hospital stay. Controls were matched by facility, year of case, sex and age. Finally, in study 3, we conducted a matched case-control study to estimate readmission rates at 30- and 90-days for mucormycosis patients compared to control patients. Results: The prevalence of mucormycosis-related hospitalizations was estimated as 0.12 per 100,000 discharges during January 2000 to June 2018. The highest prevalence and number of cases occurred in western states (i.e. Alaska, Arizona, California, Colorado, Hawaii, Idaho, Montana, Nevada, New Mexico, Oregon, Utah, Washington, Wyoming). We also found a higher prevalence of mucormycosis among patients with coagulopathy, chronic heart failure, weight loss and cardiac arrythmias. This analysis confirmed prior findings that mucormycosis was more common among patients with diabetes mellitus, hematological malignancies, and fluid and electrolyte disorders. Mucormycosis was associated with longer inpatient stays; the average length of stay for mucormycosis patients was 23 days compared to 6 days for controls matched by facility, year, sex and age. Regression analyses found that mucormycosis was a significant predictor of increased length of hospital stay, adding almost two days on average compared to matched controls. Mucormycosis patients also had a higher rate of readmission than control patients; they had 30-day readmission rates 35 times higher and 90-day readmission rates more than four times that of controls. Conclusions: While mucormycosis is not a common infection documented in US patients, it has a significant impact of patient length of stay and hospital readmission rates. The study provides an estimate of the prevalence and burden of mucormycosis among US hospital patients. The significant clinical and patient burden associated with mucormycosis showcases the importance of surveillance and understanding required to further optimize treatment protocols and protect susceptible US patients.Item Exploring grandmother kinship caregivers' perceptions of caregiving experiences(2025) Guhin, Taylor Anne; Berkel, LaVerne A.Grandmothers stepping in as kinship care providers is a growing subset of foster care. This study provided new insight into the changes a child endures when receiving care from their grandmother through Bronfenbrenner’s ecological lens. Conceptualizing child development through Bronfenbrenner’s ecological model helps explain how human development can be influenced by a child’s constantly changing immediate and larger social systems (Bronfenbrenner, 1977). Our study reviews the foster care system; including relative or nonrelative care, which can be formal or informal, as well as congregate care. There are many strengths highlighted in kinship care. For instance, in previous research, children shared that kinship care provided a protective environment that supported emotional recovery and helped them cope with adverse life circumstances (Burgess et al., 2010; Geen, 2004). Also, kinship care provides children with permanence and stability within their microsystem. However, many obstacles arise during this type of caregiving placement. Specifically, kinship caregivers are more likely to be older and have fewer economic resources (Stein et al., 2014), have received fewer educational services (Sakai et al., 2011), and to be in worse physical health than non-kin caregivers (Liao & White, 2014). Therefore, it is imperative to learn more from the grandmother’s perspective, specifically the caregiving struggles and how this change has direct and indirect effects on the child’s microsystem, mesosystem, exosystem, and macrosystem. Through a qualitative study, the research team interviewed 12 grandmothers and used Braun and Clarke’s (2006) thematic analysis guide to derive themes and subthemes from the data. This study found three main themes: challenges in caregiving, strengths in caregiving, and resources needed. Specific challenges discussed by the grandmothers were navigating relationships with biological parents, the lack of financial assistance, and the need for childcare. Some unique strengths found were that stability and safety were maintained in the child’s microsystem. Overall, this study offered insight into the child’s functioning in grandmother kinship care through the grandmother's perspective, while utilizing Bronfenbrenner’s ecological lens to examine each level of the child’s ecological system. This study's findings gave specific ideas on how to support grandmother kinship families in clinical settings and identified future research directions.Item Dynamic graph neural network framework for real-time multi-modal data analysis and predictive modeling(2025) Almousa, Ghadah Fahad M.; Lee, Yugyung, 1960-In recent years, Graph Neural Networks (GNNs) have become increasingly prominent for analyzing complex, interconnected data across fields such as transportation, social networks, and cybersecurity. Despite their advancements, many existing GNN models struggle to capture the intricate interactions among temporal, spatial, and domain-specific knowledge, particularly as these factors evolve dynamically, while also accounting for the complexities of multi-modal data in real time, with current GNN architectures often falling short in leveraging cross-modal correlations. We present a novel Dynamic Graph Neural Networks (DGNNs) Framework that integrates Partial Differential Equations (PDEs), temporal-spatial modeling, and domain-specific knowledge to address these gaps. By enabling real-time processing of multi-modal data, this framework bridges real-world dynamic systems with the evolving landscape of AI and machine learning applications. This interdisciplinary approach uniquely advances AI, machine learning, and big data analytics by harmonizing spatial-temporal dynamics, domain customization, and multi-modality integration in a cohesive framework. GNNs have become essential tools for analyzing complex, interconnected data in domains such as transportation, social networks, and cybersecurity. However, current GNN models often struggle to effectively capture the dynamic interactions of temporal, spatial, and domain-specific knowledge, especially when processing multi-modal data in real time. This dissertation presents the development of a DGNNs Framework designed to overcome these challenges, illustrated through extensive use cases. For instance, in traffic prediction, experiments using datasets such as Performance Measurement System Bay Area (PEMS-BAY), Metropolitan Traffic Los Angeles (METR-LA), and other PeMS Performance Measurement System datasets demonstrate the framework’s superior performance in prediction accuracy and robustness, effectively managing real-world data variability and spatial-temporal dependencies. Additionally, the framework efficiently models inter-variable dependencies in Multivariate Time Series (MTS) forecasting in domains such as energy, weather, and environmental monitoring, achieving stable long-horizon predictions through its PDE-enhanced graph structure. Ultimately, the framework’s capabilities extend to social media analysis for misinformation detection and rumor spread pattern discovery, with superior classification results on datasets like Pheme, Twitter15, Twitter16, and WEIBO. These examples showcase how the framework uncovers evolving patterns across platforms by processing multi-modal data inputs such as text and network interactions, surpassing traditional models.Item Data-driven modeling of infections using real-time location data and electronic health records(2025) Thota, Ravi Chandra; Uddin, Md Yusuf Sarwar (Mohammad Yusuf Sarwar)Hospital-acquired infections (HAIs) and respiratory disease outbreaks continue to pose significant challenges in indoor environments where the transmission dynamics are usually influenced by human movement patterns and contact tracing analytics. Traditional infection simulation models are based on synthetic movement assumptions and uses limited spatial representations which limits their capacity to analyze real-world disease transmission patterns. This dissertation addresses these limitations using data-driven modeling approach based on real-time movement and clinical data. The study outlines three enhancing disease/ infection modeling methodologies (i) mathematical modeling, (ii) agent-based simulation, and (iii) discrete-event simulation which are applied to two different contexts (i) academic indoor gatherings and (ii) hospital settings. Firstly, Ultra-Wideband (UWB) Real-Time Location System (RTLS) devices were deployed on four academic gatherings to capture second-level indoor movement data. These trajectories were used to model contact-driven respiratory disease spread dynamics in indoor settings through a mathematical modeling approach to calculate contact metrics, estimate transmission rates, and simulate basic reproduction numbers. Then, UWB RTLS devices were deployed in a post-surgery observation unit of an urban hospital to collect healthcare workers and mobile medical devices movement data. An agent-based model to simulate healthcare-associated infection transmission risk was developed to evaluate the impact of biosecurity measures such as mask use, hand hygiene, and device cleaning frequency, enabling systematic assessment of various intervention scenarios on infection spread dynamics. Further, electronic Health Record (EHR) data from an urban hospital were utilized to develop a discrete-event simulation model of Emergency Department patient flow, addressing the lack of operational understanding in infection and risk assessments. The model was developed to quantify how patient arrivals, service procedures, and the length of stay (LOS) affects the congestion and extended occupancy in specific ED areas. By modeling patient flow with observed clinical data the approach helps to identify where patients spend the most time and where crowding takes place thereby offering insight into delays and prolonged patient presence in certain ED areas. In summary, this dissertation demonstrates that integrating RTLS-derived movement data and EHR-based clinical data into modeling frameworks increases the accuracy, interpretability and practical relevance of disease models.Item Collaborative deep reinforcement framework for multimodality integration and learning(2025) Chandrashekar, Geetha; Lee, Yugyung, 1960-; Bumann, ErinPanoramic dental radiographs are essential in modern dental diagnostics, providing a comprehensive two-dimensional view of the oral cavity to identify abnormalities, restorations, and pathologies. However, manual interpretation is prone to variability caused by human fatigue, radiographic artifacts, and overlapping anatomical structures, which can reduce diagnostic precision and consistency. Although deep learning methods have improved automated segmentation and identification, most existing systems operate as static and isolated models, limiting their ability to adapt, collaborate, and improve once deployed in clinical environments. This dissertation develops and evaluates a collaborative deep reinforcement learning framework to address these limitations by enabling adaptive cooperation among multiple learning components. The proposed framework integrates complementary perception and identification models through a reinforcement-driven collaboration mechanism that dynamically optimizes their joint predictions. Rather than relying on fixed ensemble rules, an agent learns to adjust collaboration strategies based on performance-driven rewards, improving robustness under challenging conditions such as missing teeth, restorations, and orthodontic appliances. The framework operates through three key stages: independent model learning, collaborative inference via adaptive fusion, and reinforcement-based refinement guided by policy optimization. A carefully designed reward function promotes accurate localization, boundary consistency, and anatomically valid predictions, allowing the system to reduce uncertainty and iteratively improve its performance. Experimental evaluation on a publicly available panoramic radiograph dataset demonstrates that the proposed approach consistently outperforms conventional deep learning baselines in both tooth segmentation and identification. The framework achieves segmentation and identification accuracies that exceed 98%, while significantly reducing false detections and improving F1-scores in various test scenarios. Qualitative analyzes further confirm the improved boundary delineation and numbering consistency. From a clinical perspective, this work advances the development of self-improving and autonomous dental imaging systems capable of supporting diagnostic decision-making with greater accuracy and reliability. By bridging static inference with adaptive reinforcement driven collaboration, the proposed framework offers a scalable and interpretable solution with potential for extension to three-dimensional imaging, restoration analysis, and other real-world dental and medical applications.