22689 research outputs found
Sort by
“ARE WE EVEN HELPING THESE CHILDREN?”: ETHICAL TENSIONS IN MEDICAL MISSIONS FOR BLADDER EXSTROPHY-EPISPADIAS COMPLEX
Get PDFBladder Exstrophy-Epispadias Complex (BEEC) is one of the most severe and rare congenital diseases compatible with life. The medical and mental health implications significantly impact the quality of life (QOL) of those affected. The underdeveloped and exposed bladder in BEEC leads to urinary incontinence coupled with atypical genitalia and other anatomical differences. The societal response to such physical differences often impacts the affected person with stigma, shame, and isolation. Providers who care for those with BEEC appreciate the unique needs of this population, as well as health inequities exacerbated by socioeconomic and citizenship status, race, and gender. To mitigate these health disparities, medical missions for BEEC have been implemented in low-income-countries (LIC) by pediatric urologists in the past several decades. These missions engage global experts in surgical intervention and medical services with a focus on capacity building (training and education) with local partners. There is a great deal of literature and dialogue identifying the potential harms of medical missions to the receiving community. Surgical missions are at risk for perpetuating hierarchal dynamics of oppression and creating unsustainable protocols that are not context specific. Missions that have shortcomings in cultural competence, procurement of critical products and supplies, preparation, collaboration with local providers, continuity, sustainability, and an emphasis on capacity building may produce short term solutions and/or less than optimal outcomes. This “help” from foreign providers may exacerbate a distrust of local providers perpetuating an already dire situation. Outcomes monitoring may offer insight into how the mission is fulfilling the goal of providing medical intervention in an ethical fashion. This paper will acknowledge the uniqueness of BEEC, explore ethical principles and human rights values that are especially challenged in the care of BEEC, and present recommended strategies that aim at reducing unintentional harms of medical missions
CHEMICAL ENGINEERING FOR SUSTAINABILITY: REACTIVE AND REVERSIBLE SORPTION OF NITROGEN OXIDES AND CATALYTIC UPCYCLING OF WASTE PLASTICS
Get PDFIn this dissertation, sustainable chemical engineering approaches were applied for sustainability major environmental challenges: nitrogen oxides (NOx) emissions and plastic pollution. Nitrogen oxides from combustion engines pose significant environmental issues, especially during cold-start conditions where low-temperature (98% NOx removal at 100-400 °C and nearly stoichiometric release at higher temperatures, with a reversible NOx storage capacity of 270 μmol g-1. Integrating this sorbent with ammonia-selective catalytic reduction (NH3-SCR) achieved >90% de-NOx efficiency across the exhaust process.
Addressing plastic pollution, an effective upcycling strategy was developed via catalytic hydrocracking of high-density polyethylene (HDPE), low-density polyethylene (LDPE), polypropylene (PP), and polystyrene (PS) to produce value-added aromatics such as benzene, toluene, xylene (BTX). This strategy involved plastic vaporization followed by catalytic hydrocracking using a metal-exchanged zeolite, Ir@ZSM-5. The
iii
dual functionality of the catalyst arose from the acidic support providing cracking and the metallic site facilitating hydrogenation-dehydrogenation. BTX yields exceeded 77% for PS, 70% for HDPE and LDPE, and 35% for PP hydrocracking. The catalyst demonstrated high activity, BTX selectivity, and recyclability through stability testing. Systematic control studies optimized vaporization temperatures and catalytic hydrocracking temperatures and inlet gas flow rates. Techno-economic analysis demonstrated great market potential of our developed system for catalytic upcycling of plastic waste. Life-cycle assessment provided important guidance for catalytic upcycling of plastic waste towards a circular and sustainable economy
CARDIOMYOCYTE ISOLATION AND ITS USE IN CHIP-SEQ
No full textCardiovascular disease is one of the most serious diseases all over the world which leads to a high death rate. Cardiovascular diseases are caused by a combination of genetic and environmental causes. Doxorubicin is a traditionally and frequently used chemotherapy drug for over 40 years and is a known cardiac toxin. It is used in 50% of all pediatric cancers. However, many have shown that doxorubicin can cause long-latency heart failure and the underlying mechanism is unknown. To study the toxicity of doxorubicin and determine the potential mechanisms, we developed the long-latency and acute animal (rat model) and cell culture models to mimic patients who are treated with chemotherapy and study the short-term and long-term toxicity of doxorubicin
Toward a Praxis of Culturally Proactive Family–School–Community Partnerships: Implications on Teacher Education and Development in Independent Schools
No full textFamily–school–community (FSC) partnerships are known to be impactful to student outcomes. However, though many partnership initiatives aspire to include all families, and there is an increasing body of resources for schools to do so, historically marginalized families often do not have access to these initiatives in a meaningful way. This issue is particularly salient in independent schools, given the history of exclusion in these institutions. As such, there is an opportunity for educators to examine the development and use of inclusive practices to ensure access and belonging for all families. In this dissertation, I propose a Praxis of Culturally Proactive Family–School–Community Partnerships, fusing existing research around FSC partnerships with social justice frameworks to give form to dispositions educators should espouse around FSC partnerships. Then, drawing upon the Praxis of Culturally Proactive FSC Partnerships and existing literature on effective transformational professional development to inform instructional design, I describe a participant-centered and identity-based professional development module designed for preservice and current educators to develop identity awareness, cultural proactiveness and equity literacy in FSC partnerships, and strategies to help educators act for meaningful relationships and partnerships with the families they serve
THE NEXUS OF URBANICITY, HOUSEHOLD FOOD SECURITY AND WOMEN’S DIET QUALITY IN NEPAL: A MULTI-LEVEL ASSESSMENT ACROSS YEARS AND AGROECOLOGICAL ZONES
Get PDFBackground: Urban residence has long been considered health-protective, but recent unstructured and unsupported urbanization has led to negative outcomes as well. The most commonly used urbanicity measure (rural vs. urban) fails to represent areas with both rural and urban traits, where most people live today. While household (HH) food security is a research priority in rural and urban areas, and its measurement has evolved into more sophisticated experience-based scales, links to diets are lacking. This relationship is less explored longitudinally, and across the urbanicity gradient, without consensus on the extent that urbanness can protect food security or diet quality. Objective: This dissertation aimed to develop and validate a novel measure of urbanicity in Nepal, ascertain urbanicity’s influence on HH food security across the country’s agroecological zones, and explore the temporal relationship between HH food security and Tarai women’s diet quality via dietary diversity. Methods: This dissertation utilized community, HH and individual-level data from four annual, same-season surveys collected as part of the Policy and Science for. Health, Agriculture and Nutrition (PoSHAN) study between 2013-2016 in Nepal. A nationally- and agroecological zone- representative sample within 63 wards included mixed longitudinal components. A secondary data source, the 2011 Nepal Census, provided data for one urbanicity scale item (population density). From 23 candidate variables, an urbanicity scale of 14 variables grouped within eight domains was constructed and assessed for performance using factor analysis and principal component analysis; a novel housing quality index assessed the construct validity of the urbanicity scale. Community urbanicity was treated as a predictor of HH food insecurity and as a covariate of interest in the relationship between HH food insecurity and women’s diet quality. HH food insecurity was measured using the HFIAS. Women’s dietary diversity was calculated from 24-hour recall food frequency questions. Results: Nepal’s average urbanicity score was 35.35 (out of a possible 80); the agroecological zone averages of 30.80, 36.34, and 38.62 in the mountains, hills, and Tarai, respectively, did not statistically differ (p-value > 0.05). Multi-level, mixed effects regressions showed a 10-unit community urbanicity increase significantly protected against food insecurity nationally (OR: .82; CI: .71 - .94; p-value < 0.05); stratified by agroecological zone, this relationship was significant in the mountain zone only (OR: .71; CI: .54 - .92; p-value < 0.05). Between 2014 - 2016, HH food security continuously increased, while women’s dietary diversity achievement increased then decreased. A woman had 46% lower odds of achieving a diverse diet if she resided in a moderately food insecure HH (OR: .54; CI: .41-.71; p-value < 0.05), vs. in a food secure HH. Conclusions: Increasing community urbanicity predicted lower HH food insecurity, however zones differed in effect size, and whether working in agriculture was helpful (Tarai) or harmful (hill and mountain zones). Tarai women’s diet quality was predicted by HH food security - static and temporal fluctuations – with moderate HH food insecurity posing greatest risk to women not achieving a diverse diet. Research should explore what HH and community conditions are women’s diets most compromised. Policy efforts should examine agricultural livelihoods and pathways towards greater food security, recognizing a fuller gradient of urbanicity seen such LMIC contexts
Advances in Autonomous Underwater Vehicle Technologies for Enhanced Harbor Protection
Get PDFThis Dissertation is the culmination of coursework and research focused on the challenges of, and potential solutions to, defending harbor environments against adversarial threats. Advancements are reported on in three technology areas - perception, autonomy, and vehicle fabrication - that suggest the utilization of autonomous underwater vehicles (AUVs) as a viable harbor protection system.
Chapter 2 reports a literature review documenting present-day protection systems, AUVs, and recent advancements in technology that apply to both harbor protection systems and AUVs. Chapter 3 presents a simple mathematical model of the harbor environment that is used to evaluate the effectiveness of protection solutions. An initial set of harbor protection requirements and functional objectives help define research objectives.
This Dissertation also reports on research, design, and testing of technology components that enable the use of AUVs for the harbor protection problem. Chapter 4 presents a study on the application of convolutional neural network-based algorithms onto commercially available low size, weight, and power electronics to confirm that a forward looking sonar is able to detect and classify multiple divers in real time. Chapter 5 reports an autonomy architecture, as well as underlying modes and behaviors, that demonstrate how an AUV may search, detect, classify, and deny threats within a harbor environment. Chapter 6 presents a study of additive manufacturing processes and their utilization for design of novel bulkheads, custom pressure vessel structures, and a vehicle hull form.
The reported hull form, with integrated hardware and autonomy, constitutes a prototype AUV that was designed and fabricated at the Johns Hopkins University Applied Physics Laboratory as part of this research. Testing of the prototype vehicle was completed to confirm hardware integration and collaborative autonomy capabilities. The testing that was presented in this Dissertation suggests that AUVs can support harbor protection functions.
The Appendices include four published conference papers, one patent, and one manuscript that were completed as part of this research. The Dissertation fulfills, in part, the requirements for the Doctor of Engineering, and serves as a collection of documents that can inform future harbor protection research and technology development
NOVEL METHODS IMPROVE GENOME ANNOTATION
Get PDFIn the era of high-throughput sequencing, comprehensive genome annotation has become critical for understanding the functional complexities of life. Here we explore the development and application of computational methods for genome annotation, focusing on human transcriptome analysis, prokaryotic gene prediction, and circular RNA annotation. The primary contributions of this work span three distinct areas, each addressing unique challenges in the field. First, we develop a structure-guided isoform identification approach that utilizes three-dimensional protein structure predictions to identify functional human gene isoforms. Our method evaluates over 230,000 isoforms of human protein-coding genes assembled from thousands of RNA sequencing experiments across various human tissues. We identify hundreds of isoforms with more confidently predicted structure and potentially superior function compared to canonical isoforms, thus demonstrating the potential of protein structure prediction as a powerful tool for genome annotation and transcriptome analysis. Second, we present a universal protein model for prokaryotic gene prediction, Balrog, which employs a temporal convolutional network to analyze amino acid sequences from a diverse set of microbial genomes. Balrog eliminates the need for genome-specific training and matches or outperforms existing state-of-the-art gene finding tools. Lastly, we introduce alignment-free methods for annotating circular RNA in humans, leveraging a simple k-mer-based data structure
INTEGRATION OF TIME AND LIGHT TO PATTERN DAILY ACTIVITY IN DROSOPHILA MELANOGASTER
Get PDFAs the Earth rotates about its axis and creates 24-hour rhythms of light and darkness across the planet, organisms adjust their physiology and behavior to align with these daily changes in their environment. Optimal adaptation to such daily, repetitive changes involve predicting and reacting to those changes properly. Seminal studies over the past few decades have shown that the genes that control circadian rhythms form robust, 24hr molecular cycles that help organisms internally keep track of time. However, behavior is not always gradual and slow; it is often rapid and discrete. It remains unclear how animals integrate their internal sense of timing with their external sense of their environment in order to properly shape their behavior. In this thesis, I will study this problem in the fruit fly, Drosophila melanogaster, where the genes and neurons that control daily output of rhythmic behavior have been the subject of intense study. Drosophila have the unique advantage of pairing remarkably powerful genetics with robust circadian behavior, allowing us to study the basis of their behavior down to the level of specific genes within individual neurons in the brain.
In Chapter 1, I will introduce circadian rhythms and their study in Drosophila. In Drosophila, a group of approximately 150 neurons called the “clock network” is responsible for generating circadian behavior. I will describe how these neurons have been studied in the past, including how they interact with each other and how they sense light. This introduction will establish the base of knowledge required to understand the significance of the work in this thesis.
In Chapter 2, I will discuss how Drosophila anticipate the light-to-dark transition at dusk. I found that E2 evening cells of the clock network are responsible for controlling evening anticipation. Further, I show that this function depends on their internal molecular clocks, but not their electrical or synaptic activity. The work in this chapter describes a specific neural and molecular mechanism through which a well-described circadian behavior is executed.
In Chapter 3, I will show that E1 evening cells of the Drosophila clock network sensitize the animal to the light-to-dark transition at dusk, allowing them to rapidly transition from peak arousal before dusk to sleep after dusk. E1 neurons accomplish this goal specifically through their electrical activity in a light-dependent manner. Surprisingly, the molecular clocks of these neurons appear to be completely dispensable for transitioning the animal from arousal to sleep, and the neural activity of E1 neurons does not change between dawn and dusk. The experiments in this chapter illustrate a novel mechanism for Drosophila to sense and react to the light-to-dark transition at dusk.
Overall, the work described in this thesis presents a model for how Drosophila pattern their activity around dusk using their internal sense of timing and their external perception of light in the environment. Molecular clocks within E2 neurons help the animal predict when dusk will occur and increase the animal’s activity leading into the transition. In addition, E1 neurons sensitize the animal to the light transition at dusk, promoting activity during daylight, but facilitating sleep onset after nightfall. Together, these neurons shape innate, daily rhythms of activity in Drosophila by allowing animals to integrate their internal sense of timing with their external sense of the environment
TOWARDS ROBUST COMPUTER VISION USING SYNTHETIC DATA AND GENERATIVE MODELS
No full textEnabling computers to understand visual signals is the long-standing goal of researchers in computer vision. Despite the huge progress made in the past decade by deep neural networks, computer vision models are still not ready for many real-world applications due to insufficient robustness to out-of-distribution (OOD) scenarios. Their good performance on standard benchmarks does not seem to transfer well to deployment environments. The need for large amounts of training data and the black-box nature of discriminative deep networks are two major challenges for robustness.
In this dissertation, I propose to pursue robust vision in two directions: 1) leveraging synthetic data and 2) generative modeling. In the first part of this dissertation, we first demonstrate how synthetic data is beneficial for evaluating model robustness. Then, we propose a synthetic augmentation framework for training discriminative models that improves the robustness to OOD nuisance factors. We identify the importance of proper domain adaptation in this framework. We further find synthetic data augmentation reduces the amount of annotated real data needed, which is desirable for tasks where annotated real data is expensive to obtain. Furthermore, we explore synthetic data augmentation at intermediate level. We propose a model that performs reasoning at a domain-invariant representation space so that no domain adaptation is needed. In the second part, generative models are discussed for model robustness. We first tackle the failure and anomaly detection problem in discriminative models with conditional deep generative models. Finally, we present an approximate analysis-by-synthesis approach with strong occlusion robustness. We build a 3D-aware feature-level generative model with robust likelihood that supports efficient inference. Putting together, this work points towards several promising directions for robust computer vision
Mass Spectrometry of the Atmosphere of Venus
Get PDFMass spectrometry is a powerful tool for studying planetary atmospheres. In this dissertation, we explore how mass spectrometry has been and can be used to understand the chemistry of the atmosphere of Venus by analyzing existing data and by creating synthetic data of future observations. The first chapter of this dissertation details the mass spectral deconvolution algorithm used in the subsequent chapters. This method is based on the work of Gautier et al. (2020) and Serigano et al. (2020) but required several changes in order to be used to model the Venusian atmosphere. The second chapter is a reanalysis of data from the Pioneer Venus mission using this updated model. We were able to verify many of the previously published results from the Large Probe Neutral Mass Spectrometer (LNMS) as well as make new measurements of isotope ratios and abundances for trace gases. We present the first ever measurements of the 32S/33S/34S, 17O/16O, and 21Ne/20Ne isotope ratios in the Venus atmosphere, and we tentatively detect Kr and Xe and their isotope ratios. We also place upper limits on numerous hydrocarbons, phosphorus- and sulfur-bearing species, and volatile metals. The final chapter investigates the capabilities of unit resolution mass spectrometers to study the Venusian atmosphere. Future missions such as NASA’s DAVINCI are expected to carry unit resolution mass spectrometers in order to measure the abundances and isotope ratios of key trace species and noble gases. Unfortunately the low resolution of unit resolution mass spectrometers means that the fragment ions of some molecules are indistinguishable from one another. We use our mass spectral deconvolution algorithm to determine the range of possible mixing ratios and isotope ratios that are compatible with the data. We compare our results to previously published predictions of the performance of the Venus Mass Spectrometer aboard DAVINCI and find that our results are in line with these expectations. The work presented here is an important link between our understanding of data from previous Venus missions such as Pioneer Venus and those that are still to come, DAVINCI and beyond