22689 research outputs found
Sort by
TAILORING LIGHT-MATTER INTERACTIONS IN 1D AND 2D SEMICONDUCTORS
Get PDFNumerous exfoliation, fabrication, and gas-phase synthetic strategies have been used to prepare two-dimensional (2D) transition metal dichalcogenide (TMD) crystals and their associated devices. In order to fully exploit the unique physical properties of 2D materials, one must precisely control their dimensionality, composition, strain state, and edge structure. Although top-down approaches can yield nanoribbon-like structures, achieving these goals through a rational bottom-up synthesis strategy is a challenging but compelling prospect. To address this challenge, this thesis presents detailed optical characterization of atomically thin MoSe2 nanoribbons synthesized through a rational directed chemical growth strategy on phosphine-treated Si (001) surfaces. This work demonstrates the applicability of our substrate-mediated growth approach to different material compositions and presents how tunable surface functionalization and manipulation using PH3 and H2 gases, respectively, influence the resulting growth and width of TMD nanoribbons with atomically sharp edges.
TMD nanoribbons synthesized on designer surfaces using the unique substrate-material interactions developed in our group have enabled the exploration of anomalous optical properties. Detailed photoluminescence studies reveal systematic blue shifts in the emission energy of the anisotropic MoSe2 nanoribbons relative to their 2D monolayer counterparts. By acquiring far-field and near-field PL maps, we also observe spatial variations in the photoluminescence intensity across the MoSe2 nanoribbon crystals and investigate the origin of this variation by treating crystals with a reducing H2 environment as well as investigating in detail the crystal edge terminations and coordination environment using low-loss electron energy loss spectroscopy (EELS).
Further investigation of the role of dimensionality, defects, and heterointerfaces on the resulting physical properties is also presented, covering our advances in tailoring light-matter interactions in 1D and 2D semiconductors using lateral heterointerfaces in 1D TMD nanoribbon crystals, precursor flux influence on TMD excitonic landscape, and a particle-seeded approach to grow size monodisperse 2D crystals. This work highlights how precise control over the dimensionality, morphology, defects, and edge quality of TMD crystals can provide new avenues to tune the unprecedented excitonic features for future quantum information science and technology (QIST) applications, including the development of deterministic single-photon emitters
EXPLORING THE ASSOCIATION BETWEEN SMOKING AND DEMENTIA AMONG A MIDDLE TO OLD AGE POPULATION IN THE UNITED STATES
Get PDFABSTRACT
Background: Smoking is a well-known risk factor for both cardiovascular disease (CVD) and dementia. However, to date, few studies have examined whether CVD mediates the relationship between smoking and dementia or whether this relationship is modified by gender or race.
Methods: This study analyzed data from the 15th wave (2020–2021) of the Health and Retirement Study (HRS), including U.S. adults aged 50 and above. Participants’ self-reported information on smoking and heart disease was used. Dementia diagnosis was based on self- or proxy-report. Logistic regression models were used to estimate the total effect of smoking on dementia (without CVD adjustment), the direct effect of smoking on dementia (adjusting for CVD), and the association between smoking and CVD, controlling for covariates. Mediation analysis assessed whether CVD mediated the relationship, estimating indirect, direct, and total effects. Interaction and stratified analyses evaluated gender and race as potential effect modifiers, adjusting for CVD and covariates.
Results: Smoking was significantly associated with dementia risk, showing a total effect (95% CI: 2.76e-6, 0.00031; p < 0.05), a direct effect (95% CI: 4.28e-7, 0.000276; p < 0.05), and an indirect effect mediated through CVD (95% CI: 1.17e-6, 0.000290; p < 0.05). Stratified analysis showed a significant association between smoking and dementia among females (OR = 1.31; 95% CI: 1.06, 1.63) and White populations (OR = 1.35; 95% CI: 1.08, 1.69), whereas no statistically significant associations were observed among males (OR = 1.24; 95% CI: 0.82, 1.89), Blacks (OR = 1.04; 95% CI: 0.64, 1.69), Hispanics (OR = 1.46; 95% CI: 0.87, 2.46), or other races (OR = 1.15; 95% CI: 0.53, 2.53). Interaction analyses indicated that neither gender (p: 0.76) nor race (p: 0.68) were modifiers of the smoking-dementia relationship.
Conclusions: Smoking was associated with dementia both directly and indirectly through CVD. Stratified analyses showed significant associations between smoking and dementia only among females and White populations. However, the interaction analyses indicated that gender and race were not modifiers of the relationship. Further research with larger and more balanced samples is needed to assess potential effect modification by gender and race
COMMUNITY HEALTH WORKERS IN URBAN INDIA: EXAMINING THEIR CONTRIBUTIONS, CHALLENGES, AND POWER DYNAMICS IN HEALTH SYSTEMS
Get PDFBackground: India’s Accredited Social Health Activist (ASHA) program, the world’s largest all-female community health worker (CHW) initiative, has been extensively studied in rural contexts, but little is known about its implementation in urban settings. This dissertation explores the roles, challenges, and power dynamics of urban ASHAs in Punjab, India, through: (1) the structural constraints shaping ASHAs’ work, (2) the expansion of their responsibilities beyond core health tasks, and (3) the power asymmetries that reinforce their financial precarity and systemic surveillance.
Methods: This qualitative study was conducted in two urban sites in the state of Punjab, India. Over a three-month period, 28 ASHA workers were engaged in participant observation. Observations included door-to-door community outreach, vaccination sessions, hospital visits, survey work, ad-hoc tasks, meetings, and interactions with community members and other health workers. Additionally, 25 in-depth interviews were conducted with ASHAs and key program stakeholders, along with two community-level focus group discussions. A financial participatory session was also conducted to document ASHAs’ economic realities.
Results: Urban ASHAs play a vital role in bridging marginalized communities to healthcare, yet their work is constrained by overextended population coverage, limited supervisory support, and inadequate financial compensation. Beyond their core maternal and child health duties, ASHAs are frequently tasked with unpaid responsibilities assigned by non-health ministries while also engaging in supplementary work to sustain their livelihoods. Despite their essential contributions, ASHAs face rigid power structures that restrict their agency, reinforcing systemic control and delaying payments. Attempts to challenge these conditions are often dismissed, leaving ASHAs caught between their commitment to their communities and the limitations imposed by the health system.
Conclusion: This dissertation highlights critical gaps in the implementation of urban CHW programs, emphasizing the need for urgent structural reforms to create a more equitable and sustainable workforce. Beyond India, these findings contribute to global discussions on CHW labor rights and urban health system governance and their roles in health systems; these underscore the need for policies that prioritize CHW well-being, fair compensation, and long-term sustainability in rapidly urbanizing contexts
OPTIMIZATION OF DIFFERENT GROWTH CONDITIONS OF SELECTED BACTERIA TO ENHANCE LARVICIDAL ACTIVITY AGAINST MOSQUITO LARVAE
Get PDFThe increasing prevalence of insecticide resistance in mosquito populations presents a significant challenge to vector control programs, particularly in regions affected by arboviral diseases such as dengue, Zika, and chikungunya. This thesis explores the larvicidal potential of environmental bacterial isolates as alternative biocontrol agents targeting Aedes aegypti mosquitoes. From an initial large-scale screening, three promising isolates; Serratia marcescens 37.12, Citrobacter sp. 6.10, and Actinobacteria 10.12, were selected for detailed evaluation.
Bacterial biomass was prepared under five different culturing conditions using LB and YPD media, and tested in multiple formulations, including air-dried powders, concentrated pellets, and fractionated supernatants. Results demonstrated that larvicidal efficacy varied significantly among strains, formulations, and growth conditions. Air-dried powders derived from LB cultures generally showed higher activity than those from YPD, although performance was somewhat inconsistent, likely due to passive exposure and reduced larval ingestion. In contrast, agarose-based pellets with non-live bacterial airdried powders and fish meal as an attractant encouraged active feeding and resulted in more consistent larval mortality. Among the tested bacterial powder concentrations, 200 mg/mL pellets often outperformed 400 mg/mL pellets due to improved physical integrity and a better balance of bioactive concentration.
Further assays comparing non-live bacterial cell and culture supernatant fractions revealed that S. marcescens 37.12 retained larvicidal activity in both fractions, indicating the presence of secreted metabolites and cell-associated toxins. In contrast, Citrobacter 6.10’s activity was primarily localized in the cells, suggesting a reliance on cell-bound compounds. Actinobacteria 10.12 exhibited modest but consistent activity, with effects concentrated in the cellular fraction.
This study identifies optimized culturing conditions and practical formulation strategies for microbial larvicides, provides insights into the localization of active substances, and lays the groundwork for future studies leading to field testing. Key limitations, such as formulation stability, assay scalability, and variability in larval group size, are acknowledged. Future directions include expanded adulticidal testing, compound characterization, and semi-field validation. These findings contribute to the growing body of research supporting microbial-based approaches for sustainable mosquito control
Mixed Messages: Ukraine and the Limits of Russia and China's "No Limits" Partnership in the Information Space
Get PDFRussia’s full-scale invasion of Ukraine in February 2022 has generated intense interest in Moscow and Beijing’s so-called “no-limits” partnership, unveiled shortly before the first Russian tanks crossed the Ukrainian border. This dissertation explores the question, “what do Russia and China’s Ukraine-related messaging – both overt and covert, across different languages, and on a range of sensitive topics – reveal about the extent to which their strategic narratives converge (or diverge), and about the nature of their cooperation in the information space?” Using a dataset of nearly 1 million Twitter/X and Facebook posts in more than 25 languages over the first year of the war and analytical techniques from data science and political science, I argue that China and Russia’s strategic narratives about Ukraine diverge significantly, suggesting very different information eco-systems operating largely independently with little evidence of sustained cooperation over time. However, the degree of narrative alignment between Beijing and Moscow varied greatly by topic and whether the messenger was overtly connected to the government: In overt channels, Russian and Chinese messengers dedicated broadly the same proportion of content to criticizing the Western alliance, alleging NATO hypocrisy about the war, and addressing international policy responses such as sanctions or military assistance. However, depending on the specific issue – such as the role of Moldova in the war, or human rights abuses in Bucha – Chinese overt messengers appeared to be limited in their ability and/or willingness to echo Russia’s strategic narratives. I argue this divergence in overt narratives is driven by three apparent factors: international perceptions (and potential reputational costs for echoing Russian conspiracies), divergent strategic interests, and a relatively decentralized state media bureaucracy that allowed Chinese regional bureaus to frame Ukraine reporting in less favorable terms for Moscow. In contrast, in covert channels, Chinese and Russian narratives tended to align more strongly, and Chinese covert accounts appeared to be filling a gap in Beijing’s public strategic messaging about Ukraine, using the veneer of deniability to propagate unpopular or sensitive content that Chinese state media or diplomats were unwilling to send, presumably for fear of undermining the PRC’s publicly “neutral” position in the conflict
3D VISUAL REASONING FOR ROBUST AUTONOMOUS DRIVING SYSTEMS
No full textPerception and visual reasoning models are crucial components of autonomous driving systems. Having a good understanding of the surrounding environment enables these systems to make accurate decisions. 3D visual data is a rich source of information about the world around us, making it extremely useful for teaching machines to perform tasks such as object detection and route planning. There exist a variety of datasets curated for 3D visual reasoning, from LiDAR datasets to rich 3D shape repositories. However, leveraging this type of data in deep learning research comes with its own set of challenges. Many trained models lack generalizability and robustness to unseen data distributions or require huge amounts of labeled data to achieve the desired performance. These issues impact the use of such models in real-world scenarios. In this thesis, we focus on three research areas that address these problems. First, we address the issue of distribution shift in LiDAR datasets which adversely affect the performance of 3D object detection networks. We propose source-free domain adaptation techniques to bridge these gaps without requiring access to source data during adaptation. Second, to reduce reliance on extensive labeled datasets, we introduce a self-supervised learning framework that harnesses the geometric properties of 3D data. By encouraging equivariance under spatial and temporal transformations, this framework enables data-efficient fine-tuning of 3D object detection networks. Third, we explore the role of language in enhancing 3D scene understanding and decision-making. We propose CLIP Goes 3D (CG3D), a framework that adapts vision-language models for zero-shot 3D shape recognition. Additionally, we introduce DiMA, an end-to-end autonomous driving system that distills knowledge from large language models into vision-based planners, achieving robust performance without the computational overhead of language models at inference time. Collectively, these contributions advance the robustness, efficiency, and semantic understanding of visual reasoning models, paving the way for more reliable and scalable autonomous systems
HEARING LOSS AND SLEEP DISTURBANCES: LINKING BOTH TO NEUROIMAGING MARKERS AND COGNITIVE DECLINE
No full textModification of single risk factors for Alzheimer's Disease and Related Dementias (ADRD) prevention has had limited success. Multidomain intervention might better address the multifactorial nature of ADRD and common co-occurrence of risk factors. Hearing loss and sleep disturbances are two promising targets that remain overlooked in public health recommendations. This dissertation aims to investigate the associations between hearing loss (audiometric hearing, speech-in-noise performance) and sleep disturbances (self-reported duration and quality) in aim 1, and to examine the combined associations of hearing loss and sleep disturbances with global and domain-specific (language, memory, attention, executive function, visuospatial ability) cognitive decline in aim 2 and neuroimaging markers (brain volume, β-amyloid deposition, white matter hyperintensity, white matter microstructural integrity, functional connectivity) in aim 3. Aims 1 and 2 used data from 802 participants aged ≥50 years in the Baltimore Longitudinal Study of Aging 2012-2023. Aim 3 used data from 159 participants aged ≥50 years in the Biomarkers of Cognitive Decline in Older Individuals Study 2015-2023. Aim 1 reported a unidirectional association from sleep quality to changes in audiometric hearing but a bidirectional association between sleep duration and speech-in-noise performance. Aim 2 found a faster global cognitive decline associated with hearing loss, combined with >7 or 6-7 hours of sleep. A significant interaction was suggested for executive function, where hearing loss and 6-7 hours of sleep combined had accelerated decline. Aim 3 found that hearing loss and short sleep duration might jointly contribute to smaller brain volumes and greater white matter hyperintensity, while hearing loss was individually associated with worse white matter integrity and short sleep duration was individually associated with greater β-amyloid deposition. There was a trend towards higher functional connectivity in frontoparietal control and visual networks (potential compensatory mechanisms) but lower connectivity in default mode and dorsal attention networks (deficits in cognitive processes) with both hearing loss and short sleep duration. While research gaps remain, hearing loss and sleep disturbances are two interrelated ADRD risk factors that might have unique and shared neuropathological pathways leading to cognitive decline. Future studies with objective sleep assessments, longer follow-up, and more diverse cohorts are needed
IMPLEMENTATION AND EVALUATION OF A TRAINING PROGRAM TO IMPROVE PATIENT NAVIGATORS' COMPETENCIES: A QUASI EXPERIMENT AT A PUBLIC TERTIARY HOSPITAL IN CHINA
Get PDFBackground: Patient navigation is vital for improving healthcare accessibility and patient experience in China's public hospitals, where high patient demand meets limited medical resources. Patient navigators (PNs) assist patients through the complex healthcare system, but the lack of standardized training and evaluation hampers their ability to meet patient needs. This study piloted a Competencies Improvement Training Program (CITP) in a tertiary hospital to clarify PN competencies, design a feasible curriculum, assess its efficacy, and share insights with peer hospitals.
Methods: The CITP used the Plan–Do–Check–Act (PDCA) framework and designed a curriculum with Miller's Pyramid Model. Over 6 months, eight sessions were conducted, including theory, case studies, etc. The quasi-experimental design compared PN competencies and patient satisfaction before and after. Multiple instruments measured baseline competencies and program efficacy with a 6-month post-training follow-up.
Results: A total of 75 PNs (75%) participated and completed all sessions. A total of 1189 patients were surveyed before training, 495 in the first month after training, and 502 in the 6-month follow-up. The CITP significantly boosted PN competency scores from an average of 90.259 to 95.453, though it dipped to 92.721 by 6 months. Patient satisfaction with PN services improved modestly over 6 months. Challenges in applying theoretical knowledge to practical skills were noted, suggesting differentiated training based on navigator demographics. Patient satisfaction for aspects like politeness and tone was linked to patient age and education.
Conclusions: The CITP enhanced PN core competencies and provided an evidence-based curriculum model. Future research should involve larger multi-center populations with longer-term follow-ups to validate the program's effectiveness across diverse settings
Intermediate-Temperature Membrane-less H₂/O₂ Fuel Cells
No full textHydrogen fuel cells are promising candidates for clean and efficient energy conversion in transportation and stationary applications. However, conventional proton exchange membrane fuel cells (PEMFCs) rely on polymer membranes such as Nafion™, which suffer from high cost, strict water management requirements, and thermal instability above 80 °C. In this work, we develop and evaluate a membraneless hydrogen-oxygen fuel cell system designed to operate under elevated temperatures (up to 120 °C) using a liquid electrolyte interposer. The cell employs two porous Pt/C electrodes and a circulating acid or alkaline electrolyte to conduct ions without a solid membrane. Room-temperature tests show superior performance in acidic media compared to alkaline, and Intermediate-Temperature testing reveals stable open-circuit voltage and enhanced power output due to accelerated reaction kinetics. Our results demonstrate that a simplified, thermally tolerant membraneless architecture can reduce system complexity while maintaining performance, providing a viable path toward next-generation hydrogen fuel cells
Self-reported Practice, Perceived Barriers, And Facilitators of Shared-decision-making of Chinese Dermatologists in Medical Aesthetics Encounters
Get PDFThere is wide recognition of the importance of shared decision-making (SDM) in healthcare encounters. SDM provides a way for the opinions of both patient and providers to be considered in an encounter. The provider outlines the benefits, risks, and uncertainties of various care options based upon best available evidence, while the patient provides insight into their goals, values and preferences. Ultimately the two parties reach a shared decision [1, 2]. However, there has been little research on the physician’s role in facilitating SDM in China during dermatology encounters. This dissertation is presented in a 3-manuscript format and aims to provide a comprehensive description of the use of SDM by practicing Chinese dermatologists, and analysis of characteristics of dermatologist that predict SDM implementation during the consultation and treatment process of aesthetics. The paper also identifies barriers and facilitators to SDM behavior, to provide recommendations to improve quality of decision-making, and promote patient-centered care (PCC).
Chapter 1 provides the research background, research objective and study aims, then illustrates the Integrated Behavior Model (IBM) conceptual framework, as well as the development journey of survey questionnaire. Chapter 2 is a literature review of SDM articles published in biomedical peer-reviewed journals and provides overview of the status of its application in practice, measurement tools, and barriers and facilitators of SDM. Chapter 3 presents an evaluation of Chinese dermatologists SDM practice through SDM-Q-DOC and its association with physician characteristics. Chapter 4 identifies. barriers and facilitators to SDM implementation followed by analysis of its association with physician characteristics. Finally, Chapter 5 presents a conclusion with a summary of key findings from the thesis papers and r recommendations for interventions to enhance SDM in dermatology encounters.
In conclusion, dermatologists survey reported that overall SDM is being implemented in medical aesthetics practice in China. Female dermatologists perceived more facilitators and fewer barriers and used SDM more than males. To enhance SDM among Chinese dermatologist, promotion could focus on male dermatologists, dermatologists with less working experience, and who work at public institutions. The integration of SDM into clinical practice among dermatologists could benefit both patients and dermatologists themselves