Institutional Repository

The UDSpace Institutional Repository collects and disseminates research material from the University of Delaware.

Faculty, staff, and graduate students who want their research material hosted in UDSpace should contact the University of Delaware Insitutional Repository team at openaccess@udel.edu.
Faculty may use UDSpace to fulfill the University of Delaware Faculty Senate Open Access Resolution, and in many cases may use it to fulfill open access requirements from grant funding agencies.
Departments can use UDSpace to publish or distribute their working papers, technical reports, or other research material.
UDSpace also includes all doctoral dissertations from winter 2014 forward, and all master's theses from fall 2009 forward.

To learn more about UDSpace, and how you can make your research openly accessible to the public, visit our UDSpace Information website.

Communities in DSpace

Select a community to browse its collections.

Recent Submissions

The Influence of Blood Lipids on Cerebral Perfusion by Apolipoprotein E Status
(Journal of Lipid Research, 2026-05-09) Decker, Kevin P.; Rizzi, Nick A.; Rigas, Zoe; Awad, Catherine; Habash, Elizabeth M.; Kramer, Mary K.; Cerjanic, Alexander M.; Cohen, Matthew L.; Johnson, Curtis L.; Martens, Christopher R.
Elevated blood lipids are strongly associated with increased risk of dementia due to Alzheimer's disease (AD). The Apolipoprotein E (APOE) gene plays an important role in lipid transport and is a known genetic risk factor for the development of AD, with increased risk in ε4 carriers. Reduced cerebral blood flow (CBF) is known to precede the onset of AD pathology; however, the associations between blood lipids and cerebral perfusion by APOE status are not completely understood. This study included 65 midlife and older adults (≥ 50 years old), of which 18 (28%) carried the APOEε4 allele. Using arterial spin labeling, we measured gray matter (GM) and white matter (WM) cerebral blood flow (CBF), and hippocampal blood flow (HBF). Pearson correlations and simple linear regressions were used to assess the associations between blood lipids and cerebral perfusion. Serum triglycerides (TG) and very-low density lipoprotein cholesterol (VLDL-C) were negatively associated with GM CBF, WM CBF, and HBF in all participants. However, when stratified by APOE status, the negative associations of TG and VLDL-C on cerebral perfusion were more pronounced in ε4 carriers than non-ε4 carriers, despite no significant group differences in blood lipids. High-density lipoprotein cholesterol (HDL-C) was positively associated with only WM CBF, without any differences between APOE status. No other blood lipids were associated with resting cerebral perfusion. These findings suggest that blood lipids can influence resting cerebral perfusion, and ε4 carriers are more negatively affected by this association which may partially explain the increased genetic risk for AD pathology.
Comparative hyperparameter optimization of object detection models for precision monitoring of cucumber beetles and similar insects on yellow sticky cards
(Scientific Reports, 2026-05-14) Mafuwe, Kudzai; Venkata Sai Dulam, Rohit; Kambhamettu, Chandra; Crossley, Michael S.
Computer vision presents a great opportunity for improving pest monitoring in agriculture, particularly for yellow sticky traps, a critical component in IPM. However, despite the growing interest in applying object detection models for insect identification, insect datasets present unique challenges, and approaches for fine-tuning model parameters to achieve reliable performance remain limited. This study explores the influence of fine tuning three key hyperparameters (learning rate, optimizer type and batch size) on the performance of two popular object detection models (YOLO and RT-DETR), in detecting pests on yellow sticky traps, with a particular emphasis on identifying cucumber beetles. Results showed that higher learning rates reduced performance across precision, recall, and mAP50 for both models. In contrast, SGD improved outcomes, particularly for RT-DETR, while YOLO proved more robust to high learning rates. Our study also showed that both models achieved comparable accuracy levels, once optimal settings were determined for each model. These findings highlight the importance of hyperparameter tuning for reliable pest detection systems and support the development of scalable AI workflows for precision agriculture.
When AI Rewrites the News: How Sentiment, Framing, and LLM Disclosure Shape Perceptions
(Proceedings of the 2026 CHI Conference on Human Factors in Computing Systems, 2026-04-13) Khatiwada, Prerana; Pappu, Varun; Bagozzi, Benjamin E.; Mauriello, Matthew Louis
Public concern over media-driven polarization and the rise of AI modified news has sparked interest in how sentiment and framing shape perceptions. This study examines variations in sentiment (neutral vs. extreme) and framing (balanced vs. one-sided) in LLM transformed news, along with disclosure of LLM involvement, to assess effects on readers’ emotions, perceptions, and credibility judgments. In a 2×2 between-subjects experiment (≈180 U.S. participants) plus a baseline control (45), articles were adapted from real news and transformed with LLMs. Results show extreme sentiment worsened outcomes, heightening negative emotions and lowering trustworthiness, while framing exerted more nuanced effects. Balanced news articles with extreme sentiment elicited amplified perceptions of bias and surprise consistent with the Hostile Media Effect, where balanced coverage appears biased due to amplified opposing viewpoints. Disclosure of LLM involvement modestly improved trustworthiness without undermining fairness or credibility. Overall findings highlight the need for transparent, user-facing interventions and editorial oversight in AI-mediated journalism.
AN EXAMINATION OF THE STABLE ISOTOPE DYNAMICS OF CLEARNOSE SKATE (ROSTRORAJA EGLANTERIA) TISSUES
(University of Delaware, 2023-05) Bandlow, Serena
Rostroraja eglanteria is a common, but under-researched elasmobranch species in the Delaware Bay. This study sought to elucidate the stable isotope dynamics, such as turnover rate and trophic discrimination factor, of R. eglanteria dermis, muscle, and blood plasma through a months-long diet-swap experiment with 12 fully-grown specimens in captivity. The results from this study indicate that R. eglanteria displays a remarkably slow turnover rate, as evidenced by strikingly high trophic discrimination factors ranging from 4.06 - 7.42, even after 5 months of experimentation. Potential explanations for this slow rate of isotopic incorporation include the lack of visible growth displayed by R. eglanteria in captivity, as well as the generally low metabolic rates of benthic species. These results have significant implications for future ecological studies into R. eglanteria and other similar organisms, suggesting the need for longer-term research. The methodologies presented in this study are broadly applicable to stable isotope analysis research in both laboratory and field settings, offering a viable approach to further explore the ecological role of under-studied species like R. eglanteria.
CalmSet: A Domain-Specific Test Collection for Affective Music Retrieval for Children with ASD
(Proceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval, 2026-07) Karwankar, Abhishek; Stapley, Liam; Stevens, Daniel; Mauriello, Matthew Louis
Information Retrieval (IR) increasingly relies on subjective, graded, and natural-language notions of relevance, motivating the development of reproducible test collections for ranking and recommendation. In affect-sensitive music domains, however, such resources with human-validated relevance signals remain scarce. We introduce CalmSet, a test collection for emotion-tagged music retrieval and recommendation in a therapeutic context for children with Autism Spectrum Disorder (ASD). CalmSet contains 432 modular music tracks instantiated from four purposefully composed base songs with controlled provenance, each formed by distinct combinations of seven active musical layers. Each track is annotated with ranked top-3 therapeutic intent labels and natural-language descriptions. Annotations are produced via a hybrid human-in-the-loop pipeline: CLAP proposes candidate intent labels, a large language model generates auxiliary semantic descriptions, and crowd workers provide ranked judgments without exposure to model outputs; f inal labels are aggregated using a Borda-based procedure. As initial baselines, we evaluate five one-vs-rest multi-label classifiers over CLAP audio embeddings, observing moderate micro-F1 scores (up to 0.60) but low exact-match accuracy (<0.10), while top-3 label overlap is substantially higher (Jaccard@3 up to 0.48), motivating graded-relevance evaluation. CalmSet supports both sparse (e.g., BM25) and dense audio–text retrieval models using therapeutic labels or natural-language descriptions as queries.