22689 research outputs found
Sort by
Controlling colloidal interactions, structures, and assembly in fields
No full textThe interactions of colloidal particles in a high frequency (~MHz) AC electric field needs to be understood to provide the basis for controlled assembly. The interaction between induced dipole of a single particle with the electric field is directly measured, analytically derived and empirically simplified from first principle theory. The induced dipole can also interact with a neighboring dipolar particle and together determine the assembled microstructure. Both dipolar interactions are significantly simplified and expanded to apply on more geometries. The model prediction agrees with experimental measurements of single particle interaction in field and ensemble particles assembly structures. The simple interaction model allows fast equilibrium and dynamic simulations, enabling machine learning tools to find optimal assembly policy.
By balancing the local osmotic pressure with dipolar interactions, a model was built to predict equilibrium density profiles and assembly states. The resulting simple analytical model is shown to accurately predict the equilibrium structures for a broad range of particle shapes, system sizes, and field conditions.
Finally, feedback controlled self-assembly of elliptical and rectangular particles between multiple stable, metastable and transient states with different positional and orientational order are tested with high accuracy, fast response and large system size. An equilibrium guided proportional controller is designed and enables migration-assisted assembly and disassembly that achieves fast reconfiguration response. Control policy can make informed decisions to rescue the trapped assembly state, escape the kinetic barriers by morphing the free energy landscape. A defect-free large single 4-fold colloidal crystal can be obtained on the scale of 10 minutes can be obtained, which otherwise cannot be achieved within reasonable timescale
SINGLE MOLECULE PROBING OF MISMATCH TYPE AND SEQUENCE DEPENDENCE OF DNA MISMATCH REPAIR BY MUTS
Get PDFExtracellular and intracellular processes can modify the genomic DNA, where errors, including base-pair mismatches or base insertion-deletion loops (IDLs) can be introduced. Accumulation of these errors lead to mutations, and if left unrepaired, can cause disastrous outcomes, such as diseases and cancers in humans. Luckily, organisms possess a DNA mismatch repair pathway that scans, recognizes, and repairs these errors. MutS, one of the first proteins to initiate the DNA mismatch repair pathway, scans the DNA until it recognizes and binds to a DNA mismatch. ADP to ATP exchange makes MutS adopt a conformation called a sliding clamp, where MutS releases the mismatch and slides along the DNA to recruit other repair proteins. Previous in vivo studies suggest that the E. coli mismatch repair efficiency is hypervariable and depends on the mismatch type and local sequence context. Deletion studies suggested that MutS is responsible for the observed hypervariability. However, the molecular mechanism of variable repairing efficiency for different mismatches was not well understood. In this thesis, I used single molecule FRET (fluorescence resonance energy transfer) to directly detect MutS binding to a mismatch, sliding clamp formation, and dissociation.
In Chapter 2, I examine the mechanism for MutS recognizing different DNA mismatches and single IDLs in vitro. The effective rate of sliding clamp formation in the presence of ATP was correlated with in vivo repair efficiencies. The in vitro binding behaviors for single base insertion-deletion loops were similar to that observed for highly repaired mismatches. In Chapter 3, I characterized the kinetic proofreading mechanism of MutS during sequence-dependent mismatch repair. It was shown that MutS recognizes single IDLs better than longer IDLs. MutS regulates the initiation of DNA mismatch repair at two checkpoints: when MutS initially contacts a DNA mismatch and when MutS exchanges ADP for ATP molecules. Only after MutS recognizes and confirms the presence of an error in the DNA, it signals other repair proteins. Overall, these studies support a model where MutS regulates the initiation of the mismatch repair pathway
Giving Away Liberalism: Philanthropy, Technologists, and the Making of a New Establishment
No full textBy 2008, a new generation of billionaire tech executives and investors had orchestrated a seeming revolution in American philanthropy. They developed a new approach which used innovation and entrepreneurship to solve the world’s toughest problems. But theirs was merely the most recent of modern philanthropy’s supposedly revolutionary moments across the late twentieth century. Throughout each crisis and its aftermath, liberal foundation leaders refined their ideas on encouraging entrepreneurial individuals to address social problems in their own communities. They saw entrepreneurship as a means of challenging racial discrimination and expanding opportunities to participate in democracy. Faced with strengthening political and economic headwinds, however, they gradually came to accept social innovation in the nonprofit sector as their best chance to make some, if limited, progress toward their vision of society.
This dissertation reveals this longer history of entrepreneurship as a vehicle for change through the ideas of an influential and ever-widening network of liberal leaders in philanthropy, public policy, the social sciences, business, and activism. It reconstructs the crucial role that this network played in developing entrepreneurship as a vehicle for achieving change, shifting the center of philanthropy from New York City and Washington, DC to the West Coast, altering the relationship between philanthropy and the state, and diversifying the nation’s establishment and its elite institutions. Giving Away Liberalism contemplates the possibilities, and the limits, of philanthropists’ partiality for innovation as the path to achieving transformative social change
Teaching at the Intersection: Examining the Wellness and Retention of Black Women Educators in K-12 Settings
No full textThis dissertation examines the wellness and retention of Black women educators within
kindergarten to twelfth-grade educational institutions across the United States. Despite Black
women educators' critical role in fostering inclusive and supportive learning environments, they
often face unique challenges due to their intersecting gender and race identities that impact their
professional sustainability. Through a mixed-methods approach, combining qualitative and
quantitative data, this researcher identifies key factors influencing Black women educators’ well-
being and retention. Participants, drawn from various regions in the United States, shared
experiences that highlight systemic inequities, the burden of racial and gendered expectations,
and the lack of institutional support. To address these issues, this study proposes creating and
implementing affinity groups explicitly tailored for Black women educators. These groups are
designed to provide a safe space for professional development, emotional support, and collective
empowerment to support their wellness. Additionally, a complementary podcast series is
recommended as a resource to further enhance the connectivity and outreach of these affinity
groups, fostering a broader sense of community and shared knowledge. The findings and
recommendations presented in this study aim to inform educational leaders and policymakers on
the importance of targeted support structures such as affinity groups, ultimately contributing to
the retention and holistic wellness of Black women educators in K-12 settings. Additionally, the
author stresses that while affinity groups can provide essential support, they are not a substitute
for the systemic change needed to address the root causes of inequity in educational institutions
INVESTIGATING THE EMOTIONAL WELL-BEING OF HIGH-ACHIEVING, AFFLUENT ADOLESCENTS
Get PDFUsing Boyer’s Scholarship Model, this doctoral work presents scholarship of integration, scholarship of discovery, and scholarship of application. Scholarships of integration and discovery, two interconnected pieces, inform a stand-alone scholarship of application – an infographic –. The problem examined is the emotional well-being of high-achieving, affluent adolescents. Adolescents are facing an unprecedented mental health crisis worldwide, including within affluent regions of the US. As a result, educators and schools are faced with the challenge of supporting students’ emotional well-being more than ever. For high-achieving, affluent adolescents, the barriers impacting their emotional well-being are internal and external conflicts unique to their experiences. Importantly, research on this specific population’s emotional well-being remains understudied, though growing. The researcher used Bronfenbrenner’s Ecological Systems Theory (EST) to delve into the high-achieving, affluent adolescents’ intricate world to examine their emotional well-being within a school context. This exploration reveals factors that shape their emotional well-being, and the intricate interplay among them underscores the academic pressure and stress they experience. Using a convergent mixed-methods design, the emotional well-being of high-achieving, affluent adolescents are analyzed using pre-existing data from 2015 of students (N=1,460) and 2023 semi-structured interviews with mental health providers (n = 4). Results indicate high academic stress and pressure levels, with the need and urgency to address and support emotional well-being in high-achieving, affluent adolescents. Informed by the EST-guided literature review, a convergent parallel mixed-methods design aims to contribute to the understanding of the emotional well-being of this population and provide insights for practitioners
Innovative Approaches in Quantitative Phase Imaging and Raman Spectroscopy for Enhanced Biomedical Analysis
Get PDFAdvancements in imaging technologies have significantly transformed biological research, allowing scientists to explore cellular structures and processes with unprecedented detail. This thesis explores the evolution and application of two prominent label-free imaging modalities: Quantitative Phase Imaging (QPI) and Raman Spectroscopy (RS). QPI provides high-resolution, quantitative data on cellular morphology by measuring phase shifts of light passing through transparent samples. Despite its advantages, including label-free imaging and reduced photobleaching, QPI faces challenges related to analysis pipeline development. This work addresses these challenges through the introduction of CellSNAP, a fast algorithm for 3D cell segmentation, and Cell-TIMP, a framework for cellular trajectory inference based on morphological parameters. On the other hand, Raman Spectroscopy, known for its ability to reveal molecular compositions through vibrational spectroscopy, suffers from limitations such as weak signal intensity and shallow penetration depth. To overcome these limitations, this thesis presents a self-assembly-based Raman probe for enhanced detection and Spatially Offset Raman Spectroscopy (SORS) for improved depth penetration and non-invasive diagnostics of cartilage health. By developing and applying these methods, this thesis aims to advance the field of label-free imaging, providing new tools and insights for both fundamental biological research and clinical applications
Multiphase Flows in Supersonic Plume Surface Interaction
No full textMore than 50 years after the last crewed lunar landing, humans are planning to return to the Moon to establish a long-term presence. Within this century, we may also witness the first humans setting foot on Mars. To land on these planetary surfaces, spacecraft use rocket engines firing towards the surface to decelerate before touchdown. The interaction between the supersonic exhaust plume and the granular surface generates a crater that can destabilize the lander. Particles are ejected outward at high speeds, posing risks to the vehicle and nearby infrastructure. In Martian environments, ejected particles can be dispersed vertically, leading to potential damage to the vehicle and expensive instruments. These particles can also modify the high-speed flows as they are entrained, exchanging momentum with the surrounding gas through drag forces. In lunar environments, due to the much lower ambient pressure, ejecta particles tend to leave the surface at shallow angles, reaching speeds akin to the Moon's escape velocity, posing risks to nearby infrastructure. To understand these complex problems relevant to extraterrestrial landings, this thesis aims to establish a comprehensive experimental framework to identify different multiphase flow physics and unveil the key mechanisms underlying the plume surface interaction (PSI) problem.
The first part of this thesis focuses on studying gas-particle and particle-gas interactions in the nearfield region of a particle-laden underexpanded jet, simulating particle entrainment during PSI. Using the JHU Jet experimental facility, particles and gas-phase structures were visualized with schlieren imaging. Additionally, ultra-high-speed particle image velocimetry was used to measure the gas-phase properties. By examining the Mach disk movement due to the presence of particles, results were compared with numerical simulations conducted by our collaborators, leading to the development of a model that predicts the Mach disk shift caused by the particles.
The second part investigates the ejecta dynamics generated by a supersonic jet impinging on a granular bed under Martian and lunar-relevant conditions. Experiments were performed at NASA Marshall Space Flight Center inside a vacuum chamber with a diameter of 4.5 m. Ejecta particles were visualized using a high-speed camera, with particle positions and velocities obtained via Lagrangian particle tracking. A simplified Eulerian two-fluid model is presented, effectively predicting the temporal evolution of the ejecta by using measured ejecta particle concentrations and existing drag laws.
The final part of this thesis explores the mechanism of particle entrainment when a supersonic jet impinges on a granular bed in a reduced pressure environment. From the crewed Apollo missions of the 1960s to the recent uncrewed missions by the Chang'E spacecraft, a striking pattern appears in all landing videos: a ray system of ejecta streams expelled radially at high speeds. This thesis attributes the streak generation to the universal Görtler instability in the curved shear layer of the supersonic jet, which entrains particles into filaments and ejects them from the landing site.
This research provides new insights into ejecta dynamics in PSI, highlighting the crucial role of fluid mechanics and two-phase flow during planetary landings. These findings will contribute to mitigating ejecta-related risks in future missions to other planetary bodies
MEDICAL UNCERTAINTY AND CLINCIAL RESEARCH PARTICIPANTS’ RESPONSES TO NEGATIVE UNINFORMATIVE GENOME SEQUENCING RESULTS
No full textBackground: Exome and genome sequencing are increasingly implemented in clinical and research settings, yet the diagnostic yield of genomic sequencing remains low, and most patients receive negative results. Due to their complexity and ambiguity, negative results may add uncertainty about the cause of one’s condition and the utility of genomic information. Because perceptions of uncertainty may be associated with one’s likelihood to adopt genetic technology, there is a need to understand patient responses to negative uninformative results and assess the use of educational materials in returning negative results and mitigating misinterpretations or adverse reactions.
Objective: This study aimed to assess perceptions of uncertainty and engagement with genomic information in clinical research participants with negative uninformative genomic sequencing results.
Methods: Two-hundred ninety-seven clinical research participants enrolled at the National Institutes of Health (NIH) clinical center completed a quantitative survey measuring genomic knowledge, perceived uncertainty, selected personal traits, and including some open-ended questions. Participants also read an informational page briefly describing their results.
Results: Higher perceptions of uncertainty in genomics are associated with decreased likelihood to communicate results, more negative attitudes towards genetic testing, and less confidence in seeking genetic information. Most participants understood their negative results to be reassuring, uninformative for their current care, and possibly limited based on the current understanding of their condition’s genetic etiology. Many participants had questions about what was and was not included in their genetic testing, and confusion regarding their clinical symptoms and lack of genetic findings.
Conclusions: Patients with high perceptions of uncertainty may be more likely to misinterpret their negative results and be less equipped to seek further information; the opportunity to speak with a genetics provider about the meaning of the results and the possibility of reanalysis should be made readily available. Educational materials may be a useful tool for returning negative results, particularly in delineating the scope of the testing completed and the limitations inherent to genomic testing
Towards Robust Medical Imaging Diagnosis Systems via Integrating Domain Knowledge
No full textThis dissertation presents a comprehensive exploration into the integration of domain knowledge to fortify the robustness of medical imaging diagnosis systems within the realm of deep learning. As deep learning revolutionizes medical diagnostics in healthcare, it faces ongoing challenges. A significant issue is the vulnerability of medical data to noise and perturbations. This dissertation posits that the strategic incorporation of domain-specific insights stands as a pivotal solution to enhancing system robustness.
This dissertation explores how integrating domain knowledge enhances the resilience of medical deep learning projects against noise and perturbations. It investigates innovative strategies such as data synthesis to combat data annotation scarcity, subjective learning for tackling annotation noise, and employing a mixture of experts (MoE) with data fusion to address diverse data styles and computational challenges. Highlighting key contributions, the research introduces AI-driven synthetic data generation for early pancreatic cancer detection, subjective learning models for interpreting data with inherent subjectivity, straightforward result fusion for ultrasound diagnosis, and video segmentation with spatial and temporal information. Additionally, it showcases the use of MoE and fusion strategies in liver fibrosis analysis, underscoring the pivotal role of domain knowledge in improving diagnostic precision and system dependability.
In summary, this work contributes to the field of medical imaging diagnostics by showcasing how the integration of domain knowledge can mitigate challenges posed by data imperfections, leading to the development of more robust, accurate, and reliable diagnostic systems. It paves the way for future advancements in medical deep learning, aiming to enhance patient care through technological innovation and strategic insight
Restructuring America's Justice System: Punitive Policies, Mass Incarceration, Capital Punishment Contradictions, and Requisite Criminal Justice Reform
Get PDFThis thesis assesses the need for criminal justice reform and a restructuring of the American justice system. The first contribution of this thesis is an examination of the punitive nature of the United States criminal justice system, as compared to the less punitive measures of the European model, focusing on alternative prison models used in France. In assessing the programs and policies in effect to promote alternatives to incarceration and rehabilitation in France, it is evident that although the United States has initiated programs aimed at mirroring the European model, barriers to change have hindered the United States from adopting these less punitive practices found in France.
The second contribution of this thesis is an analysis of capital punishment in the United States through a regional approach, employing broad case studies of representative states to evaluate the factors and policies that have led states to adopt or reject the death penalty. Through this research, geographic disparities regarding the utilization of capital punishment were revealed. Despite inequities and concerns surrounding this method of punishment, capital punishment continues to be pursued in a number of jurisdictions, even though a majority of counties no longer support the death penalty.
The third contribution of this thesis is a data-driven approach, utilizing state-specific case studies, to display facts and statistics regarding counties within states that apply capital punishment, as a means of illustrating the geographic disparities surrounding capital punishment both among and within states. This leads to an investigation concerning the implications of states taking different approaches to capital punishment. In this study, the approach taken by Congress regarding capital punishment was explored by assessing the involvement and/or lack thereof of Congress regarding death penalty legislation. Following this, reform legislation proposed by lawmakers and officials was explored. Finally, recommendations aimed at decreasing punitive practices of the United States legal system, reducing mass incarceration, and diminishing disparities surrounding the utilization of capital punishment are presented for policymakers and leaders. An analysis of these proposed policies is conducted to determine the probability of success, as well as rationalizations for which the suggested strategies may not be feasible