19304 research outputs found
Sort by
A Seat at the Table: A Qualitative Analysis of the Socialization of Black Women in Medical School
No full textBackground: Despite efforts to diversify the physician population in the United States, Black women continue to represent less than 3% of the physician demographic despite making up 13% of the total female population in the country. They also continue to face unique challenges in medical school due to the intersection of their multiple marginalized identities. Purpose: This qualitative study explored the socialization experiences of Black women in United States medical schools within the past 15 years and how they perceived those experiences influenced their academic and professional success. Methods: Using qualitative inquiry, this study included semi-structured interviews with seven Black women who have graduated from U.S. medical schools within the past 15 years. The data collection focused on their social experiences and their perceived belonging regarding their intersecting racial and gender identities. Results: Findings of this study reveal that their social experiences were shaped by mentorship, peer support, and campus climate. Participants also highlighted how courage, resilience, and self-advocacy played a pivotal role in their academic and professional success. Findings also revealed that while in medical school, participants described feelings of isolation and exclusion as they were often one of the few Black women in these spaces. This study emphasizes the importance of inclusive and supportive structures and policies to increase the persistence and success of Black women in medical school. Conclusion: Findings reveal that mentorship, social networks, and institutional support contribute to Black women’s academic and professional success in medical school. The results of this study emphasize the need for intentional support through funding, mentorship programs, and medical education policy changes
A Refreshing Breath: Beethoven Sonata for Piano and Violin No. 10 in G major, Op. 96 and Its Style
No full textBeethoven composed a total of ten violin sonatas. The first nine were written over a relatively brief period and exhibit a clear developmental trajectory—from freshness to profundity in expression, and from agility to complexity in technique. Among them, Sonata No. 9 in A major, Op. 47 (“Kreutzer”) is widely regarded as one of the most technically demanding and grand works in the violin repertoire. After completing this monumental work, Beethoven ceased composing in the genre for nearly a decade. When he finally returned to it with Sonata No. 10 in G major, Op. 96, the style was markedly different from its predecessors—neither virtuosic nor epic. Owing to the towering status of the Ninth Sonata, the Tenth has long remained in its shadow: It has received limited scholarly attention and is rarely performed on today’s concert stage. This essay explores why the Tenth Sonata differs so greatly from its predecessors, examining the work through multiple lenses, including historical context, key figures, and compositional process. Ultimately, it reveals the sonata’s expressive depth and its essentially pastoral character. This essay seeks to contribute to the appreciation and scholarly understanding of this underrated yet invaluable gem in Beethoven’s violin sonata repertoire
Immune Dysfunctions in Neurodegenerative Diseases
No full textParkinson's disease (PD) is a progressive neurodegenerative disorder characterized by the selective loss of dopaminergic neurons in the substantia nigra (SN) and α-synuclein pathology. While neuroinflammation is implicated in PD, the role of adaptive immunity remains poorly understood. This study investigates how genetic and immune factors drive neurodegeneration, focusing on the role of adaptive immunity, particularly T-cell dysfunction, in PD progression. Here, we demonstrate the PD-associated leucine-rich repeat kinase 2 (LRRK2) G2019S mutation alters T-cell differentiation and function. Using a novel T-cell receptor (TCR) transgenic murine model, we found the mutation skews CD4+ T-cells toward Th2 while suppressing Th9 and regulatory T cells (Tregs) via Janus kinase/signal transducer and activator of transcription 3 (JAK/STAT3) hyperactivation, reversible by STAT3 inhibition. Moreover, transcriptomic analysis of patient-derived induced pluripotent stem cells (iPSCs) revealed that LRRK2 modulates antigen presentation pathways, with LRRK2 inhibition increasing major histocompatibility complex class I (MHC-I) expression and CD8+ T-cell cytotoxicity. These results demonstrate LRRK2's dual role in PD pathogenesis, initially promoting neuronal dysfunction through intrinsic mechanisms but subsequently protecting against immune damage by modulating antigen presentation, suggesting a reevaluation of LRRK2 inhibition as a treatment approach. Additionally, to further explore immune dysregulation in PD pathogenesis, we utilized autoimmune-prone mice as a representative model of hyperactive immunity. These mice exhibited dopaminergic neuron loss, microglial activation, and elevated neuron-specific autoantibodies, mirroring PD-like pathology. Inhibition of the colony-stimulating factor 1 receptor (CSF-1R), which is crucial for the development and function of microglia, reduced autoantibody levels and microglial activation; however, it failed to rescue neuronal loss. This underscores the role of adaptive immune dysfunction in driving neurodegeneration in PD. In summary, our findings provide direct evidence to link dysfunctions in adaptive immunity and PD pathogenesis from three related but independent pathways: (1) LRRK2 G2019S directly reprograms T-cell function through STAT3 signaling, (2) LRRK2 plays a bidirectional role in PD pathogenesis, initially promoting neuronal dysfunction and later protecting against immune-mediated damage through the modulation of antigen presentation, and (3) Hyperactive adaptive immune cells can promote neurodegenerative processes. These results position PD at the critical intersection of neurodegenerative and immune-mediated mechanisms, with T cells emerging as key mediators of disease progression
AI-Driven Vehicle Re-Identification: Decoding Effective Features through Ablation Studies
No full textIn today’s increasingly urbanized world, effective traffic monitoring is essential to improve road safety, manage congestion, and support law enforcement. One of the critical challenges in traffic monitoring is vehicle re-identification (ReID), which involves recognizing the same vehicle at different locations, under varying viewpoints, and across diverse environmental conditions. Humans are generally able to identify vehicles based on unique features such as shape, color, and distinctive markings. However, translating this ability into an automated system poses significant challenges due to variations in vehicle appearance caused by changes in lighting, viewing angle, and background. While deep learning approaches have achieved notable success in improving overall re-identification accuracy, limited research has focused on understanding how individual and combined vehicle attributes—such as color, type, and vehicle ID—contribute to the vehicle re-identification process, which is crucial for improving model performance. In this work, we aim to gain deeper insights into the comparative strengths of these attributes in both Convolutional Neural Networks (CNNs) and Transformer-based models. Specifically, we conduct a comprehensive ablation study using ResNet-50 and ViT-Base (16-patch) architectures, training both models in parallel, each with auxiliary branches designed to learn attribute-specific features that support the vehicle ReID task. Our findings offer interpretable insights into how these model architectures respond to different vehicle attributes. Vehicle ID consistently played a central role in enhancing the re-identification process, with further gains observed when combined with vehicle type. In contrast, relying solely on color or type was less effective—likely due to class imbalance and subtle inter-class similarities among vehicle images. Evaluations on the VeRi-776 and VERI-Wild datasets, compared with state-of-the-art results, and visual analysis of our experiment outcomes, underscore the need for better learning of fine-grained visual cues—such as lights, grills, and logos—through targeted attention mechanisms and improved cross-domain generalization to build robust and deployable vehicle ReID systems
Integrated Transcriptomic and Machine Learning Approaches for PTSD Biomarker Identification and Classification
No full textPosttraumatic stress disorder (PTSD) is a condition marked by intricate molecular and biological complexities. The objective of this research is to unravel candidate biomarkers associated with PTSD through transcriptomic analysis of the existing datasets in the public domain comparing gene expression profiles between Caucasian and Black control and PTSD-recovered individuals and Caucasians not recovered individuals. Results from our differential gene expression analysis revealed 607, 441, and 164 significant genes in Black control versus recovered, Caucasian control versus not recovered, and Caucasians control versus recovered, respectively, with nine common differentially expressed genes, including MMP8, CEACAM6, DEFA4, and DEFA1B. These genes were identified as hub genes. Functional enrichment and pathway analysis highlighted their role in immune response and extracellular modeling. Further, a range of machine learning techniques— such as Random Forest, SVM, kNN, GLMNet, and XGBoost—were utilized to classify PTSD status, with GLMNET and SVM being the best-performing models, validated using an external dataset (GSE125216). In conclusion, the findings generated through our study will provide a deeper understanding of the molecular processes underlying PTSD and offer novel diagnostic indicators and possible avenues for therapeutic intervention
Effects of Sub-grid Scale Parameterizations on Hurricane Simulations over Ocean and Urban Areas
No full textHurricanes are among the costliest natural disasters in the United States and regularly inflict severe damage on urban infrastructure. Accurate forecasts are therefore essential for preparedness and limiting these extreme events' economic toll. Numerical weather‑prediction (NWP) models—such as the Weather Research and Forecasting (WRF) system—are powerful forecasting tools. However, some of their physical parameterizations were neither designed for nor tested with real hurricanes. This thesis addresses that gap by evaluating two key parameterizations in WRF: (i) subgrid‑scale (SGS) turbulence schemes and (ii) surface‑roughness and urban canopy treatments. The first part of the study investigates how SGS eddy‑viscosity choices affect hurricane intensity, turbulence, and wind profiles. Large‑eddy simulations (LES) of five major hurricanes were run with a 1.5‑order, three‑dimensional turbulent‑kinetic‑energy (TKE) SGS scheme. Each storm was simulated under three eddy‑viscosity settings— default, halved, and doubled—yielding 15 cases. A parallel set of 10 cases employed an alternative nonlinear backscatter and anisotropy (NBA) SGS scheme. Two idealized LES runs and one fine-grid (~80 m) nested simulation brought the total to 33. Reducing SGS stress intensified storms by raising boundary‑layer wind speeds and lowering the altitude of peak winds, improving surface‑wind forecasts by ~9 % and minimum sea‑level pressure by ~29 % relative to the default setting. These results reveal that standard SGS models are overly dissipative because they overlook the rotational suppression of turbulence, underscoring the need for SGS schemes tailored to hurricane dynamics. The second part assesses how aerodynamic roughness length (z0) and urban‑canopy schemes shape near‑surface winds over cities. For four land‑falling hurricanes affecting Houston and New Orleans, increasing z0 in the Single‑Layer Urban Canopy Model (SLUCM) reduced modeled wind speeds and cut mean absolute error (MAE) by ~20 %, whereas decreasing z0 introduced large positive biases. Additional experiments compared three urban options—Bulk (no‑urban), SLUCM, and the multi‑layer Building Energy Model (BEM). The Bulk scheme delivered the most accurate surface‑wind forecasts in every nested domain, while SLUCM slightly outperformed BEM in the limited vertical‑profile data. These findings highlight the need to recalibrate urban schemes and surface‑drag parameters when applying WRF to hurricane‑force winds
Effect of Exercise Intervention on Vascular Dysfunction in Different Pathologies
No full textThe endothelium plays a key role in vascular health by activating endothelial nitric oxide synthase (eNOS). Exercise interventions (EX) can alleviate endothelial dysfunction. We previously determined that EX improves endothelial dysfunction in the aorta and coronary artery in mice with a high-fat diet (HF) and the cerebral artery in mice with Alzheimer’s disease (AD). However, the mechanisms by which EX improves endothelial function in different arteries in each disease are largely unknown, especially regarding the adiponectin and serotonin (5-HT) receptors. Therefore, this dissertation aimed to evaluate the possible mechanisms by which EX affects eNOS activity and, thus, endothelial function, in obesity and AD. I hypothesized that EX would protect eNOS activity in these conditions by improving different mechanisms. The first study investigated the effect of EX on mesenteric arterial endothelial dysfunction focusing on eNOS and its underlying mechanisms in HF mice. I found that HF reduces p-eNOS, adiponectin, and 5HT1B receptor expression, but these changes were prevented with EX. Additionally, EX reduced markers of oxidative stress and inflammation but not endoplasmic reticulum (ER) stress in HF mice. Our results suggest that EX can prevent obesity-induced mesenteric arterial endothelial dysfunction and has a broad effect on attenuating changes to adiponectin and 5-HT1B receptors, oxidative stress, and inflammation. The second study evaluated the effect of EX on aortic endothelial dysfunction focusing on eNOS and its underlying mechanisms in AD mice. I found that reduced p-eNOS in AD mice aorta was restored by EX. Notably, EX reversed elevated 5-HT1B receptor expression in AD aorta. Additionally, EX increased P2Y2 receptor expression and reduced markers of oxidative stress, inflammation, and ER stress in AD aorta. To demonstrate a role for the 5-HT1B receptor in modulating eNOS, I cultured endothelial cells with Aβ, which lowered p-eNOS similar to our AD samples. 5-HT1B receptor inhibition recovered eNOS activity, suggesting EX may mitigate endothelial dysfunction in AD by reducing 5-HT1B receptor expression. In conclusion, eNOS activation is impaired in the arteries with obesity and AD. Importantly, EX can increase eNOS activation with conserved mechanisms including the modulation of adiponectin and 5HT1B receptor expression, oxidative stress, and inflammation
Theoretical and Computational Approaches to Minimal Mobile Sensing
No full textMulti-agent systems are finding increasing applications in our world, and so the study of sensor networks, which serve as the backbone of such systems, is of growing significance. These fields are expansive and multidisciplinary with many problems on which one could focus their efforts. In this dissertation, we study mobile sensor networks, specifically the problem of evasion paths in minimal sensing mobile sensor networks, through topological tools and computational approaches. We first extend the existing bi-directional results on the existence of evasion paths to three new systems: fenceless connected, fenced power-on, and ultimately fenceless power-on, thereby establishing results that can be used in conjunction with the existing corpus for arbitrary networks encountered in application. We then introduce a new Python package that we developed for simulating sensor networks at larger scales. Additionally, we analyze the limitations of current software and algorithms in feasibly simulating large-scale networks over many trials. Subsequently, using our package, we run simulations for various power-down experiments. We present and discuss the results, which suggest that: (1) noisy radii present in real networks may aid in detecting evaders; (2) the characteristics of the varying radii are important; and (3) swarms may improve evader detection for a fixed coverage area—each under specific conditions. Finally, we construct a framework motivated by the Reeb graph from the evasion paths problem to establish results that help address a previously unexplored issue that may arise in practice and pave the way for a new family of algorithms that could significantly increase the scales at which mobile sensor networks can be simulated
Analytical and Experimental Investigations on the Transfer and Development Lengths of High-Strength Duplex Stainless Steel (HSSS) Strands in Prestressed Concrete Girders
No full textConventional Carbon Steel (CS) strands have long been used as longitudinal prestressed reinforcement in bridge elements. Although prior research has been conducted on stainless steel strands, studies specifically investigating their development length in prestressed applications remain limited. This study aims to address this gap by evaluating the transfer and development lengths of High-Strength Stainless Steel (HSSS) strands compared to CS strands, particularly in prestressed concrete girders and piles designed for corrosive environments. A comprehensive research project has been conducted on predicting the transfer and development length of CS strands, leading to various analytical and empirical models. However, the applicability of these models to HSSS strands remains uncertain due to differences in material properties and bond behavior. This study employs a comprehensive experimental program using rectangular prestressed concrete prisms to compare the transfer and development lengths of HSSS and CS strands. In an effort to enhance accuracy and reliability, a redundant experimental approach is used, incorporating multiple measurement techniques to capture strand stresses during testing. This ensures a more precise evaluation of the bond behavior and load transfer mechanisms between CS and HSSS strands. By analyzing key influencing factors identified in existing models for CS strands, this research evaluates their suitability for HSSS strands and explores potential modifications to enhance predictive accuracy. The findings provide critical insights into the bond characteristics and load transfer mechanisms of HSSS strands, contributing to improved design guidelines for prestressed concrete elements. Using a comprehensive testing methodology strengthens the reliability of this study, offering engineers a holistic understanding of HSSS strand behavior and supporting their broader application in bridge structures exposed to harsh environments
A Study of Latin* Immigrant Family Engagement at an Elementary Campus
No full textBackground: The Latin* population continues to increase, and their academic achievement remains lower than non-minority populations. The achievement gap leads to inequities in educational success for minority students. Effective parent engagement contributes to increasing academic achievement for Latin* students. However, there are barriers present that prevent effective parent engagement in the Latin* community. Identifying and addressing the barriers supports creating academic equity for Latin* students. Purpose: The goals of this study were (a) to provide additional research to the literature addressing family engagement of emergent bilingual Latin* students at the elementary level, (b) to identify specific barriers limiting the family engagement of the Latin* population at an elementary campus, and (c) to identify perceptions of Latin* family engagement from the perspectives of Latin* parents. Method: This study employed a qualitative research methodology to explore Latin family engagement at an elementary school. Individual interviews gathered data from the perspectives of Latin* Spanish-speaking parents of emergent bilingual students. The study analyzed the extent to which Latin* parents of emergent bilingual students believe family engagement is accessible. It examined Latin parent perspectives on family engagement and identified their perceived strengths and barriers to family engagement. Results: Findings revealed that Latin* parents want to engage in their child’s school and education. However, they face challenges, including scheduling conflicts and language barriers. Additionally, inconsistencies in parent-teacher communication resulted in communication being a leading factor for parent disengagement. Conclusion: The results suggest approaches for school and district leaders that can help improve Latin* family engagement. An increase in family engagement could result in higher student achievement