22689 research outputs found
Sort by
Component Separation for Cosmic Microwave Background Observations
Get PDFMany upcoming cosmic microwave background (CMB) experiments are searching for a B-mode polarization signature in the CMB. A detection of this signal would provide confirmation of the theory of inflation, which posits that the universe underwent an exponential expansion in its earliest moments. B-modes are expected to arise due to gravitational waves generated during inflation, and would thus provide compelling evidence for quantum gravity as well. Next-generation CMB experiments are poised to place constraints on primordial B-modes so stringent as to either make a detection or rule out a subset of inflationary models.
The primary obstacle facing next-generation CMB experiments is the characterization and removal of astrophysical foregrounds. The primordial B-mode polarization signal is expected to be extremely faint, and B-modes can also be sourced by foregrounds such as synchrotron emission and thermal dust. As CMB experiments become increasingly sensitive, our ability to isolate the inflationary signal will be limited by our ability to accurately model and subtract foregrounds.
The focus of this work is on modeling dust. The exact composition and emission mechanisms of interstellar dust are still unknown. Variations in the temperature and spectral index are expected to produce complex spectral shapes as a result of spatial averaging. There is also a component associated with thermal dust emission known as anomalous microwave emission (AME), which despite being discovered over 25 years ago has still not been linked to any known constituent of the interstellar medium.
In this work, I assess the ability of a moment expansion technique to fit dust with an unknown parameter distribution for an experiment such as the Probe of Inflation and Cosmic Origins (PICO). I demonstrate that this approach can eliminate bias in measurements of the tensor-to-scalar ratio while also shedding light on the dust physical properties. I also investigate AME and its possible connection to polycyclic aromatic hydrocarbons (PAHs). To do this, I create a full-sky map of PAH emission at 3.3 um using archival data from the COBE Diffuse Infrared Background Experiment (DIRBE) instrument. This map enables direct comparison with Planck measurements of AME over a variety of astrophysical environments
Gamifying Radiology Education: A Virtual Desk for Studying MRI Pulse Sequences
Get PDFMRI physics, a challenging topic for radiology residents to master, is a key component of the American Board of Radiology (ABR) CORE exam required for board certification in diagnostic and interventional radiology. A previous needs assessment identified a desire for multimedia learning resources about MRI physics among radiology residents. This thesis project investigates whether the use of interactive and gamified materials facilitates learning and engagement when studying MRI physics, with a focus on MRI pulse sequences.
We created an interactive, gamified study “desk” containing activities focused on the basics of MRI pulse sequences including a written module, a puzzle activity, a practice test, and flashcards. The module was created iteratively with a radiology attending. User testing was conducted with radiology residents studying for the CORE exam. We created a survey to comparatively test learning and engagement using our written module and the interactive pulse sequences puzzle. This online Qualtrics study was distributed to radiology residents and fellows at the Johns Hopkins University School of Medicine. Participants were sorted into either module or module + puzzle survey groups where they completed a pre-test, viewed relevant WebGL content that we created, answered an engagement survey, and completed a post-test.
Survey results showed that participants in both groups significantly improved their scores, suggesting the overall efficacy of our educational materials. Results from this study illuminate a preference among radiology residents for visual, interactive media and may inform the creation of future radiology education materials.
There was no significant difference in learning between test groups. Engagement survey results showed that the module and puzzle were equally engaging and yielded positive impressions. Respondents appreciated the module’s diagrams and graphics as well as the opportunity to build a pulse sequence using the puzzle activity. Suggestions for improvement included adding more pulse sequence puzzles and practice questions at the end of each module subtopic.
Our study suggests that interactive learning media may increase learner engagement. Future directions include increasing survey sample size, implementing respondent feedback into the WebGL content, and hosting the study tool on TeamRads.com, a publicly available radiology education website
Radiomic Estimation of Geometric Metrics of Bone Microstructure: Comparison of Ultra-High and Normal-Resolution CT
Get PDFIn recent times, there has been a range of multi-detector computed tomography (MDCT) systems introduced with ultra-high resolution (UHR) capabilities. They include the use of energy-integrating and photon-counting detectors. Our research focuses on whether texture features obtained from UHR CT scans can estimate geometric parameters of trabecular microarchitecture with better accuracy when compared to conventional normal resolution (NR) CT.
Using UHR MDCT (0.25 mm slice thickness), NR MDCT (0.5 mm slice thickness), and micro-CT (50 µm voxel size), we captured six human femora images and extracted 129 spherical ROIs (10 mm) aligned spatially via rigid registration. Twenty-four Gray Level Co-occurrence Matrix (GLCM) features from CT ROIs and four geometric metrics from micro-CT ROIs were calculated and used to train a cross-validated linear regression (LASSO) model. The model predicted geometric metrics from GLCM metrics and was evaluated using coefficient of determination (R2).
The average prediction accuracy using UHR GLCM features was 91.3% and using NR GLCM features was 52.35%. The top two predictors observed in most cases were Sum Average and Difference Average. However, in NR, they became strongly correlated possibly due to the reduced spatial resolution of NR causing a loss of contrast.
Texture features from UHR CT seem to have better predictive capabilities of the trabecular microstructure when compared to NR CT. This could potentially help improve bone health assessment capabilities
An Evaluation of the Health and Physical Development of Adolescents Living with Perinatal HIV Infection in Rural Zambia
Get PDFBackground: With the success of antiretroviral treatment, children living with HIV are aging into adolescence, which presents unique challenges. Despite being at-risk for treatment non-adherence, there are still significant gaps in knowledge about puberty, sexual activity, and treatment outcomes in this age group. This study aimed to evaluate puberty, sexual health, and HIV care and treatment outcomes among adolescents living with HIV in rural Zambia.
Methods: This analysis included two components nested within the Pediatric Antiretroviral Therapy (PART) study, an open clinical cohort study conducted from 2007 to 2023 at Macha Hospital. The first component was a cross-sectional analysis of puberty and sexual health using data from 2015-2022 among PART participants aged 12-20. Differences by age and sex were assessed using Wilcoxon rank-sum, chi-squared, or Fisher’s exact tests. The second component analyzed HIV treatment outcomes using data from 2015-2019 among PART study participants aged 5-20. Differences in retention in care by age were assessed using a linear mixed effects model (LME). Differences in viral suppression and adherence by age were assessed using logistic regression generalized estimating equation models.
Results: 168 adolescents (55% female) participated in the study on puberty and sexual health. Girls exhibited earlier sexual maturation than boys (median age of sexual debut: 15 vs. 18
years; p=0.07). Condom use with last partner among girls (35.7%) and HIV status disclosure with last partner for both girls (54%) and boys (40%) were low. In the analysis of HIV treatment
outcomes, 587 participants (50% female) contributing 6,024 clinic visits (median: 3 per year) were included. Despite a significantly lower mean number of visits per year (12-14y: -0.20, p<0.01; 15-17y and 18-20y: -0.32, p<0.01), adolescents demonstrated higher ART adherence
(12-14y: adjusted odds ratio [aOR]=1.21, p=0.07; 15-17y: aOR=2.13, p<0.01; 18-20y: aOR=2.24, p<0.01), but similar viral suppression (12-14y: aOR=0.62, p=0.19; 15-17y:
aOR=0.56, p=0.31; 18-20y: aOR=0.80, p=0.76) compared to children 5-11 years.
Conclusion: These findings emphasize the need for comprehensive sexual education, improved access to healthcare services, and tailored interventions, such as increased social
support and case management, to address the unique challenges faced by adolescents living with HIV, ultimately improving their long-term health outcomes
A View from Anomalous Branes on Generalized Symmetries in Quantum Field Theory
Get PDFThe notion of symmetries in quantum field theory has undergone various revolutionary generalizations throughout the past decade. We explore in this dissertation with explicit examples how one can characterize higher-form symmetries and non-invertible symmetries from the perspective of branes in string/M-theory. Specifically, we make use of the anomalous topological actions of both the bulk supergravity and the brane worldvolume to study the global symmetry structure of the corresponding theories.
The dissertation is divided into three main parts. In the first part, we analyze the global symmetries and anomalies of 4d field theories that arise from a stack of M5-branes probing a class of flux backgrounds. These backgrounds consist of a resolved singularity fibered over a smooth Riemann surface, supported by a non-trivial -flux configuration labeled by a collection of flux quanta. We lay the groundwork for a systematic analysis of such backgrounds by adopting anomaly inflow methods to identify the global symmetries of the 4d field theories.
In the second part, we compare the 4d anomaly polynomials obtained respectively from reducing the anomaly polynomial of parent 6d superconformal field theories on a Riemann surface, and directly from eleven dimensions using the setup considered in the first part of the dissertation. We provide a geometric characterization of a class of modes that decouple along the renormalization group flow from six to four dimensions as operators originating from M2-branes wrapping resolution cycles.
In the third part, we study the symmetry topological field theory in holography associated with 4d supersymmetric Yang-Mills theory with gauge algebra . From this, all the bulk symmetry operators are computed and matched to various D-brane configurations. The fusion algebra of the operators emerges from brane dynamics. In particular, we show that the symmetry operators are purely determined from the center-of-mass modes of the branes. We identify the fusion coefficients with the relative motion of the branes, and also establish the origin of condensation defects, arising from fusion of non-invertible operators, as the consequence of tachyon condensation in brane-anti-brane pairs
Total Domination Ratios of Archimedean Lattices
Get PDFAn Archimedean lattice is an infinite graph corresponding to the vertices and edges of a regular tiling which is vertex transitive.
A subset of vertices dominates a graph if all elements of either belong to or are adjacent to an element of .
A dominating set of vertices is a total dominating set if every vertex is adjacent to an element of the set.
A total dominating set of vertices is efficient if every vertex is adjacent to a unique element of the set.
The total domination ratio is the smallest possible proportion of vertices in a total dominating set.
We prove that a lower bound for the total domination ratio of an Archimedean lattice is determined by its inverse regularity, and such a proportion is achievable if and only if the lattice has an efficient total dominating set.
We also show that seven of the eleven Archimedean lattices have efficient total dominating sets, thereby yielding their total domination ratios.
Conversely, we show that the other four lattices cannot have efficient total dominating sets.
For two of these lattices, we determine their total domination ratios by way of forcing arguments concerning their local graph structure.
For the remaining two lattices, we provide upper bounds on their total domination ratios with explicit constructions of total dominating sets and investigate lower bounds via structural and computational analysis involving the external boundaries of subsets of possible total dominating sets
INTERACTION OF LINEAR PERTURBATIONS WITH A SHOCK WAVE
Get PDFSmall-amplitude fluctuations in compressible flows can be decomposed into three modes: vortical, acoustic, and entropic. When any of these linear waves are incident onto and interact with a shock, all three modes of fluctuations can be generated downstream. Using linear analysis and Snell's law, the fluctuations downstream of the shock are expressed in terms of the orientation, amplitude, and properties of an incident wave and the baseflow. In addition to the analytical expression, we performed Navier-Stokes simulations of elementary waves interacting with a shock. We examined the effects of incidence angle, amplitude, and frequency of incoming waves. The interaction results in unsteady shock movement and higher kinetic energy downstream of the shock. For oblique incidence to a normal shock, the generation and transmission of acoustic waves are dependent on the incidence angle and baseflow properties, but are independent of the frequency of the wave. When working with an ideal gas, the transmitted oblique acoustic wave has its pressure increased but its kinetic energy decreased. In addition, the generated acoustic waves resulting from vortical waves are found to be less intense compared to those generated by incident entropy waves
CHARACTERIZATION OF TIME ACTIVITY DATA (TAD) TO SUPPORT ESTIMATION OF PHYSICAL AGENTS IN THE ENVIRONMENT
Get PDFTime Activity Data (TAD) describe the timing, duration, and/or frequency of human activities and can be used to estimate exposure to physical agents in the environment. Due to activities influencing our contact with the environment, this information may allow for the derivation of more rigorous estimates of exposure to these agents. This thesis research aims to develop a systematic evidence map that characterizes TAD collection methods and primary activities/microenvironments recorded in studies collecting TAD to estimate exposure to physical agents. A rigorous search and screening process was previously used to determine a larger set of studies generating TAD for estimation of environmental agents and this work analyzes a subset of those studies. Tableau was used to analyze all of the data extracted from these studies. Most studies in this analysis estimated exposure to ultraviolet radiation or noise and primary TAD instruments used to collect this information were diaries and questionnaires. The structure of these instruments was also analyzed and revealed that most diaries have unstructured formats while questionnaires had structured formats, whether they were used to estimate ultraviolet radiation or noise. Many of the studies identified different forms of external validity as limitations in their research, thus demonstrating the importance of accounting for inter- and intra-individual variability through collecting data such as TAD
SINGLE-CELL GENOMICS AND THE MOLECULAR IMPACTS OF GENETIC VARIATION
Get PDFIdentifying the molecular effects of human genetic variation across cellular contexts is crucial for understanding the mechanisms underlying disease-associated loci, yet many contexts remain underexplored. Recent progress in single-cell data analysis offers new opportunities to accelerate this exploration: by collecting large-scale transcriptomic data at the resolution of individual cells, researchers can now derive cellular phenotypes such as cell type and differentiation stage through computational analysis, rather than laborious isolation of the context of interest, which is only feasible for a narrow subset of contexts clearly defined a priori.
In this work, we employ the single-cell analysis of in vitro model systems to study the genetic regulation of gene expression in previously inaccessible biological contexts. First, we examined regulatory dynamics over the course of directed cardiomyocyte differentiation. We then expanded to a uniquely flexible system called heterogeneous differentiating cultures, where we characterized regulatory dynamics across diverse trajectories representing each of the three germ layers.
In parallel, we addressed a key challenge, 'double dipping', that arises in this setting of derived phenotypes. Double dipping can cause existing methods to produce deceivingly confident yet misleading results. We discuss shortcomings of existing methods and propose a data thinning framework that enables valid inference with derived phenotypes under flexible distributional assumptions.
Together, this work advances the applicability of single-cell analysis to accelerate the characterization of gene regulation across diverse biological contexts, offering insight into the molecular architecture of complex traits and common diseases
Evaluating the contribution of the frmRAB operon on the in vitro activity of chlorine-based disinfectants among KPC-producing carbapenem-resistant Enterobacterales
Get PDFCarbapenem-resistant Enterobacterales (CRE) infections are a significant public health challenge associated with increasing morbidity and mortality rates globally. The primary mechanism driving carbapenem resistance among Enterobacterales is the dissemination of plasmid-mediated carbapenemase genes. In some cases, these plasmids may also harbor genes contributing to bacterial survival during exposure to certain disinfectant agents, such as chlorine-based disinfectants. In this study, we evaluated the contribution of frm gene insertions, co-localized on the blaKPC-bearing pUVA-like plasmid, on the in vitro activity of a chlorine-based disinfectant, BrulinTM BruTab 6S®. A cohort of 48 CRE isolates harboring the blaKPC-bearing pUVA-like plasmids were evaluated, including a subset (N=13) that also harbored the frm gene insertion. Carbapenemase production was confirmed by phenotypic methods. To observe the direct relationship between the frm gene insertions and the in vitro activity of BruTab 6S®, conjugation assays were performed to transfer the pUVA-like plasmid from the parent CRE isolates to Escherichia coli J53 recipient strain. Broth microdilution assays performed in triplicate were used to evaluate the effects of diluted BruTab 6S® on the growth of the CRE parent isolates and the transconjugants at room temperature (RT) and 35±2°C. At RT and 35±2°C, the parent CRE and transconjugant pairs demonstrated a decrease in bacterial growth as the available chlorine concentration increased. At both temperatures, 43.3% of the modal MICs of the transconjugant pairs carrying the pUVA-like plasmids with or without frm gene insertions were equivalent to that of E. coli J53, while 56.7% of the modal MICs differed by a single two-fold dilution of available chlorine. The findings presented in this study suggest that the frm gene insertions into the pUVA-like plasmids do not impact the in vitro activity of the chlorine-based disinfectant BruTab 6S® when E. coli J53 is used as the conjugation recipient organism. It was determined that E. coli J53 inherently harbors the frmRAB operon within its genome. Future studies should be conducted with a competent recipient organism that does not contain the frmRAB operon to directly compare frm genes and BruTab 6S® activity. These future studies may influence hospital disinfectant strategies for preventing and controlling hospital-acquired CRE infections