22689 research outputs found
Sort by
Joint Inference, Testing and Prediction for Competing Risks Data Using Multiple Endpoints
No full textCompeting risks data are commonly encountered in randomized clinical trials and observational studies. In some situations, the ending statuses of competing events have different clinical interpretations and/or are of simultaneous interest and the analysis of one single endpoint may not be sufficient to reveal the nature history of disease. In clinical trials, often more than one competing event has meaningful clinical interpretations even though the trial effects of different events could be different or even opposite to each other. Joint analysis for multiple endpoints have not been conducted in the existing literature, therefore new inferential and modelling methods need to be developed.
In this thesis, we develop joint inference and testing methods for competing risks data using multiple endpoints within restricted mean time framework. Several simulation studies and real data applications show the validity of our proposed methods. We also propose a novel two-stage prediction model, allowing us to make dynamic predictions on both outcome and survival probability. These approaches are particularly useful for better evidence synthesis and decision-making in clinical trial settings
FACE TOUCH DETECTION TO REDUCE DISEASE TRANSMISSION
No full textIn the last several years, multiple research teams have investigated the face-touch detection problem for the purpose of developing technology that can detect and subsequently prevent face-touching. IMU-based datasets have been cultivated under controlled settings of trial participants performing different face-touches. This thesis seeks to use one well-outlined dataset in particular, to build a “pre-face touch detection model” and a “completed face-touch detection model”. These two Human Activity Recognition (HAR) models are designed to be situated in a free-living study of participants “in the wild”. The pre-face touch detection model is the focus of this thesis, as we primarily seek to build a face-touch prevention tool that potentially can reduce disease transmission for its user. The need for researchers to collect free-living face-touch trials such that a pre-face touch detection model can be built that is generalizable to users “in the wild” is what motivated the development of the completed face-touch detection model. The main contribution of this thesis to the sub-problem of face-touch detection within the HAR space is our investigation of utilizing Discrete Wavelet Transform (DWT) on IMU sensor signals in order to extract localized time-frequency features relevant to the tasks of pre-face touch detection and completed face-touch detection. Additionally, we identify constraints on Precision and Specificity metrics in order to minimize the invasiveness of the pre-face touch detection model, and we maximize Recall score with these constraints considered
Is It Hard To Hide A Bear In The Desert? The Russian Intervention In Libya And The Kremlin's Quest For Great Power Renewal
No full textWhat is the character of modern Russian expeditionary warfare? I explore this question through a detailed analysis of the Russian intervention in the Libyan civil war from 2015 until mid-2020. Based on my analysis, I argue that the intervention can be best understood as an armed attempt to manage an international conflict in a similar way that Moscow manages its domestic politics.
The key characteristics of the Russian intervention in Libya were the weaponization of diplomacy, widespread overt and covert disinformation operations, and the limited deployment of military force. Moscow’s weaponization of diplomacy aimed to manipulate Libya’s domestic political system and related international peace processes toward outcomes that favored Russian interests. Russian overt information operations portrayed Moscow to global audiences as a benevolent great power that was attempting to stabilize Libya, while simultaneously denying reports that Moscow was also involved in destabilizing activities. Kremlin affiliates simultaneously conducted covert disinformation operations to support Moscow’s preferred actors in Libya, to denigrate the UN-backed government, and to criticize any actor or process that threatened Moscow’s interests. Finally, the Kremlin used covert military support to manipulate the Libyan National Army’s weaknesses and to gain Russian access to important oil infrastructure and airfields in Libyan National Army-held territory.
This characterization of the Russian intervention in Libya corresponds with four Russian ideas about modern politics and contemporary warfare. These ideas are sovereign democracy, gibridnaya voyna, the strategy of limited actions, and non-linear warfare. Sovereign democracy provides a lens through which to understand Moscow’s weaponization of diplomacy. Gibridnaya voyna provides vital perspective on information’s role in shifting a target state’s geostrategic orientation toward Russia. Finally, the strategy of limited actions combined with non-linear warfare help to understand the Kremlin’s use of limited covert military force for expeditionary operations
DEVELOPMENT OF A HUMAN iPSC DERIVED MICROGLIA ASSAY FOR EVALUATION OF TARGET DEPENDENT ACTIVATION AND PHARMACOLOGICAL INHIBITION
No full textNeurodegenerative diseases occur due to persistent chronic inflammation, resulting in
neuronal cell death. In the Central Nervous System (CNS), this inflammation is coordinated by
microglia, the resident immune cells. Inflammasomes are multiprotein complexes that are formed
in microglia in response to pathogen-associated molecular patterns (PAMPS), as well as damage-associated molecular patterns (DAMPS), and other aberrant proteins found in the CNS
environment. Pre-clinical data and patient evidence suggest that the NOD-like receptor (NLR)
family pyrin domain containing 3 (NLRP3) inflammasome is typically active in neurodegenerative
diseases. Multiple studies replicate the NLRP3 in in-vitro models to understand the disease
mechanism and pathway, and these include immortalized cell lines, primary cell lines, and
peripheral iPSC-derived lines. Current in-vitro tools used to study the NLRP3 pathway are peripheral
immune cells and are not representative of neurodegenerative images. In the present study, we
used microglia cells derived from iPSCs (iMGLs) to mimic the neuro-inflamed environment and
determine whether they are suitable in vitro models for studying NLRP3-mediated CNS
inflammation. Using IL-1β as the endpoint response, we evaluate the dose-dependent inhibitory
response of the small molecule MCC950 against NLRP3 inflammasome activation.
Correspondingly, we performed the analysis on THP1 and PBMCs (peripheral blood mononuclear
cells) to determine parallels and differences in their responses. The results of these studies
demonstrate that iPSC-derived microglia cells are efficient in vitro tools for studying NLRP3
inflammation
Oral history of E.R.
No full text“E.R.,” a member of the class of 2023, talks with Kristen Diehl about her Cuban heritage, including her parents’ journey as political refugees from Cuba to the United States in 1993 and growing up in Kendall, Florida. She describes her academic experience as a neuroscience and math double-major, including research and clinical jobs, studying abroad in Spain, and working as a TA for both math and neuroscience faculty. In addition to academics, E.R. shares her participation in campus activities, including working as the Blue Jay mascot at campus events and her involvement with the Catholic community at Hopkins. E.R. goes on to describe her goal of attending medical school and facilitating greater access to healthcare in the Latino community
Bayesian shape-constrained regression for quantifying Alzheimer disease progression with biomarkers
No full textSeveral biomarkers are hypothesized to indicate early stages of Alzheimer’s disease, well before the cognitive symptoms manifest. Their precise relations to the disease progression, however, is poorly understood. This limits our ability to diagnose the disease and intervene effectively at early stages. To better understand the relation between disease and biomarkers' progressions, we model biomarker trajectories as increasing functions of age using monotone regression splines with smoothness penalty. We further introduce two kinds of shape constraints in the regression curves to capture a scientifically-informed structure and to improve model interpretability. First, biomarker progression is believed to follow a sigmoid shaped course, in which the changes start gradually at the beginning and level off towards the end of disease progression. We capture this behavior through a “window function” that controls the prior variances on splines near the two ends of the disease progression. Second, progression curves with at most one inflection point increase interpretability of the model, since they present smoother trends of biomarker change and pinpoints the stage when the change rate becomes highest. We fit this shape-constrained monotone regression model under Bayesian framework for its straightforward uncertainty quantification and for its ability to incorporate individual-level effects. The model fit to the BIOCARD data recovers the biomarkers progressions generally consistent with the existing scientific hypotheses. The Bayesian model developed here also allow for a future hierarchical extension to multi-database settings
Arsenic Exposure and Maternal and Child Health in Rural Northern Bangladesh
No full textToday, ~45 million people in Bangladesh are exposed to arsenic in drinking water above the World Health Organization guideline value of 10 µg/L. Of these, ~20 million people are exposed above the Bangladesh national standard of 50 µg/L. Early-life arsenic exposure has been associated with adverse maternal and child health outcomes, including altered immune responses, increased risk of infections, and more. Studies reporting these associations imply that intervening to reduce arsenic exposure would improve public health, but the benefits of interventions have remained largely unquantified. The Pregnancy, Arsenic, and Immune Response (PAIR) Study is a prospective pregnancy and birth cohort in rural Gaibandha District, northern Bangladesh. The PAIR Study recruited pregnant women (n=784) in the early second trimester. These women were followed to the end of pregnancy, and, after live births, mother-infant pairs were followed to 3 months postpartum. In Chapter 1, using g-computation, I aimed to estimate the causal effects of hypothetical interventions to reduce household drinking water arsenic or urinary arsenic during pregnancy on the incidence of influenza-like illness in infants, 0-3 months of age. I found that intervening on urinary arsenic could reduce the incidence of influenza-like illness among infants in the sample by 2.3 to 19.5 cases per 1,000 person-weeks. In Chapter 2, I aimed to identify drinking water and urinary element mixtures among pregnant women at enrollment. Arsenic tended to co-occur with toxicologically significant elements in drinking water (Fe, Mn, Mo, W) and urine (Mo, W). An intervention to reduce arsenic exposure is likely to affect exposures to these elements, too. In Chapter 3, I aimed to estimate associations between anthropometric measures and arsenic methylation among pregnant women at enrollment, adjusting for biomarkers of one-carbon metabolism micronutrient status. All measures, especially those related to adiposity, were positively associated with arsenic methylation efficiency before and after adjustment. The role of arsenic methylation in the relationship between arsenic exposure and disease remains uncertain and may vary by outcome, but these results suggest adiposity via arsenic methylation may be an important source of effect heterogeneity in studies of early-life arsenic exposure and maternal and child health
On the Fringes of Slavery: White Settlers and Indentured Migrants in Nineteenth-Century Cuba
No full textThis dissertation studies colonization schemes and migratory projects in Cuba between 1817 and 1874. During this period, about 40,000 Spaniards, 2,000 Yucatecans, and 125,000 Chinese immigrants arrived in the island having signed migratory contracts that limited their freedom and set restrictive conditions on their incorporation into the labor market. Rather than focusing on intellectual and political debates surrounding them, as previous scholarship has done, this dissertation focuses on the development of these projects on the ground—on how immigrants navigated their conditions and how they shaped legislation and subsequent labor experiments. By blending imperial and social history, it reconstructs a genealogy of forms of bonded labor that developed alongside the expansion of plantation slavery and Spain’s efforts to maintain territorial control of Cuba. Such a genealogy, I contend, allows for a better understanding of the Cuban labor market as a whole and sheds light onto the way both enslaved people and bonded laborers developed connected strategies for freedom
From Dissonance to Competence: Confronting Teachers' Cultural Disequilibrium
No full textIncreasing diversity among the student population in U.S. schools has caused a widening demographic divide between teachers and students. As a result teachers’ evident lack of cultural competence, the ability recognize, appreciate, and adapt to cultural differences, has posed concerns about apparent inequitable teaching and disciplinary practices in schools. The demographic and cultural disparities between predominately White, middle class educators and students from diverse backgrounds have prompted school leaders to seek out solutions to ameliorate cultural conflict. This explanatory sequential mixed methods study used cultural historical activity theory (CHAT) to examine factors that impact urban students’ and teachers’ experiences of cultural disequilibrium as well as transformative professional learning efforts to increase teachers’ culturally responsive behavior management through restorative practices training. The findings of this project suggest that improving restorative practices professional learning by incorporating transformative learning opportunities may provide readily available tools to assist educators in navigating experiences of cultural disequilibrium
A comparison of tissue dissection techniques for diagnostic, prognostic, and theragnostic analysis of human disease
No full textHistopathology has historically been the critical technique for the diagnosis and treatment of human disease. Today, genomics, transcriptomics, and proteomics from specific cells, rather than bulk tissue, have become key to understanding underlying disease mechanisms and rendering useful diagnostic information. Extraction of desired analytes, i.e. nucleic acids or proteins, from easily accessible formalin-fixed paraffin-embedded (FFPE) tissues allows for clinically-relevant activities, such as sequencing biomarker mutations or typing amyloidogenic proteins. Genetic profiling has become routine for cancers as varied as non-small cell lung cancer and prostatic carcinoma. The five main tissue dissection techniques that have been developed thus far include: bulk scraping, manual macrodissection (MMa), manual microdissection (MMi), laser-capture microdissection (LCM), and expression microdissection (xMD). In this review, we discuss the importance of tissue dissection in clinical research, the basic methods, applications, as well as some advantages and disadvantages for each modality