Graduate Dissertations

Xinhao Wang from Nutritional Sciences presents his dissertation on an integrated strategy to reduce strawberry waste using bio-based nanocomposite coatings and a deep learning-powered quality monitoring system. This research addresses the critical challenge of postharvest strawberry loss by developing a comprehensive, dual-domain solution. The findings offer promising sustainable strategies to reduce food waste, enhance food safety, and improve quality management in the perishable food supply chain.

Understanding the Cooling Properties of Forested Wetlands and their Role as Climate Moderators   Heat has been a major issue in the United States being the deadliest weather-related event in the past 30-years. There have been several heat mitigation strategies cities have implemented such as street trees and stormwater infrastructure, but despite these implementations they have not dealt with the main symptoms of heat islands. Land cover change from natural land to dense concentrations of pavement, building and other surfaces that absorb heat contribute to the phenomenon. In contrast, wetlands are cooling islands that can regulate their local microclimate. We aimed to (1) investigate the phenomenon with a systematic review to view current research, (2) conduct our measurement campaign to understand wetland buffer-width and wetland characteristics that influence cooling, and (3) view local climate policy in Connecticut towns and how wetlands fit into their framework. Findings have found that wetlands act as a climate moderator keeping cool in summer and warm in winter. Hydrology plays a huge influence on wetland cooling. Connecticut towns could benefit greatly if they were to implement wetlands into their towns to mitigate heat and improve air quality.

Microbial communities profoundly impact host biology by influencing immune development, metabolic processes, and therapeutic outcomes. However, accurately profiling these communities at sufficient resolution to capture subtle, biologically meaningful differences remains challenging. Variability in microbiome composition—even among genetically identical laboratory animals under controlled conditions—can confound experimental results and hinder reproducibility. Addressing this challenge requires methods capable of experimentally standardizing microbiomes and computationally profiling them with strain-level resolution. First, I will present a bioinformatics workflow capable of handling large-scale microbiome datasets with high resolution. This workflow was used to demonstrate that a single antibiotic-free cecal microbiome transplantation (CMT) effectively standardizes gut microbiomes across genetically diverse mouse populations, reducing unwanted variability without promoting antibiotic-resistant pathogens. I will then introduce AmpliconHunter, a tool developed to facilitate precise identification and tracking of these microbiomes. AmpliconHunter is a highly scalable computational tool for accurate PCR amplicon sequence prediction using degenerate primers. Its performance enables efficient strain-level profiling and evaluation of primer pairs with similar accuracy and significantly improved speed when compared to existing methods. I will also detail my work creating AmpliconHunter2: a SIMD-accelerated update to the original that completes analysis for V1V9 primers on the ~2.4M genomes from the AllTheBacteria project in 38.73 minutes, compared to 419.45 minutes for AmpliconHunter (~10.8x speedup). Finally, I will introduce another tool: Microbiome HiFi Amplicon Sequence Simulator (MHASS) creates realistic synthetic PacBio HiFi amplicon sequencing datasets for microbiome studies, by integrating genome-aware abundance modeling, realistic dual-barcoding strategies, and empirically derived pass-number distributions from actual sequencing runs. MHASS generates datasets tailored for rigorous benchmarking and validation of long-read microbiome analysis workflows, including ASV clustering and taxonomic assignment. Together, these innovations will provide practical, robust and scalable methods to address microbiome variability, improving reproducibility and translational potential in microbiome-focused biomedical research. All tools are made available on GitHub under an MIT license. AmpliconHunter and AmpliconHunter2 are also made available as freely accessible web servers at https://ah1.engr.uconn.edu and https://ah2.engr.uconn.edu/.

Title: Walking Through the Sacred: How Sacredness Transforms Embodied Perceptual Experiences Program: Ecological Psychology

This is a doctoral dissertation defense in Kinesiology

Major Advisor: Grégory Pierrot Associate Advisors: Katharine Capshaw, Erika Williams Departmental Readers: Eleni Coundouriotis and Najnin Islam

Event will be held in the Library Class of 1947 Room. Virtual attendance details can be found at swapecoevo.weebly.com/phd-defense.

DNP student presenting "Leveraging AI to Improve Clinical Appropriateness of Inpatient Designation in a Utilization Management Setting."

This dissertation focused on deep learning methods for large-scale dish classification and nutrient estimation through ingredient-guided RGB-D imaging supported by vision-text contrastive learning and Swift UI iOS application development.

Efficiency-Aware Computational Intelligence for Resource-Constrained Manufacturing Toward Edge-Ready Deployment

Examining practitioners views of Sport for Development (SfD) Learning, Leadership, and Education Policy (Concentration in Sport Management)

COMMITTEE Dr. Eva Lefkowitz  Dr. Keith Bellizzi  Dr. Amanda Denes

Stellar evolution and the synthesis of the elements are governed by key nuclear reactions, among which the fusion of 12C with an alpha particle to form 16O, denoted as the 12C(α,γ)16O reaction, is "of paramount importance". The ratio of carbon to oxygen produced during stellar helium burning, which is determined by the 12C(α,γ)16O reaction, allow us for example to predict the fate of massive stars, whether they end up as neutron stars or black holes. Despite five decades of study, this reaction's cross section remains poorly constrained at the astrophysically relevant energies. This thesis presents the development and implementation of a new method to measure the cross-section of the 12C(α,γ)16O reaction by measuring the time-reverse process – the 16O(γ,α)12C reaction – using a Time Projection Chamber (TPC) operated in intense γ-ray beams. The first-generation optical readout TPC (O–TPC) was constructed at UConn and used at the High Intensity γ source (HIγS) facility at Duke University. Building on these results, a next-generation electronic readout TPC (eTPC) was constructed and commissioned at the University of Warsaw, incorporating a fully digital electronic readout system for high-rate data acquisition. The eTPC was exposed to quasi-monoenergetic γ-rays from 8.51–13.9 MeV, corresponding to Ecm=1.4-4.8 MeV of the 12C(α,γ)16O reaction. A comprehensive analysis framework was developed to identify the 16O(γ,α)12C events and reconstruct their kinematics. This permitted angular distributions of the photo-dissociation events to be examined. The analyzed angular distributions yield results which are consistent with a fundamental prediction of quantum mechanics, a feat not seen in earlier data sets. The results demonstrate that this method can achieve accurate event reconstruction, clean event separation, accurate energy calibration, and angular resolution sufficient for astrophysical studies. This work establishes the validity of our new method for precision measurement of the 12C(α,γ)16O reaction through its time reverse process. This paves the way toward future measurements at lower energies with reduced uncertainty and improved extrapolation to stellar conditions.

This dissertation focused on the design and development for fluorescent sensor array for the foodborne pathogenic bacterial and biofilm identification with machine learning techniques. it also includes the investigation of interfacial biofilm monitoring and quantification for better pathogenic biofilm control and food safety.

Time projection chamber (TPC) detectors, operating in gamma beams, were used to study the structure of 12C and 16O at energies above 10 MeV. In this energy region, the algebraic cluster model (ACM), which models 12C as an equilateral triangle of alpha particles, predicts new states in 12C, which are relevant for the structure of excitations of the Hoyle state (E = 7.6542 MeV). The Hoyle state is responsible for the production of 12C in the universe, and its structure is central to stellar evolution. Specifically, the current experiment searched for a predicted 2+ state at around 12 MeV, belonging to a rotational band corresponding to the exotic bending-mode vibration. The structure of 16O is directly relevant for oxygen formation during stellar helium burning. An electronic TPC detector (eTPC) was prepared at the University of Warsaw with assistance from UConn students, when possible during the pandemic. It was used at the High-Intensity γ-Source (HIγS) at Duke University. The HIγS facility is ideally suited for ℓ = 1 (E1) and ℓ = 2 (E2) nuclear excitations. Data were collected at gamma energies ranging from 8.51 to 13.9 MeV, with low background and substantial statistics. The high-energy results, above 10 MeV, were the focus of this thesis. Algorithms were developed to analyze tracks recorded in the eTPC. It records in three dimensions, with high efficiency and low background, all relevant kinematic variables characterizing the reactions. These yield a complete, detailed view of the interaction. Complete and detailed angular distributions of 12C and 16O photodissociation spanning the entire angular range (0 to 180◦) were measured and analyzed using a partial wave decomposition. The results are compared with fundamental predictions of quantum mechanics, i.e. unitarity, and to specific nuclear structure models. No new states were found for 12C below 12.5 MeV. The 16O data improve our ability to extrapolate measurements made at laboratory energies to stellar conditions.

Social Work Dissertation Defense (Ph.D.) will be held virtually.

DNP student presenting "The Healing Power of Movement: Integrating an Evidence Based, Nurse Driven Mobility Assessment Tool as the Established Gold Standard on an Orthopedic Surgical Unit."

DNP student presenting "Self-Monitored Blood Pressure Devices for Hypertension Management in Undeserved Communities."