22689 research outputs found
Sort by
Product Design Features, Packaging, and Online Retailer Marketing of Cannabis Products Containing Cannabinoids Other than THC and CBD: Opportunities for Cannabis Regulation
Full text linkBackground
Cannabis product characteristics affect product safety, and communication of these characteristics through product packaging and retailers informs consumers about potential risks, benefits, and psychoactive effects of cannabis products. Little is known about product characteristics, labeling, and marketing of cannabis products containing cannabinoids other than delta-9 THC and CBD, which remain unregulated in many jurisdictions. This dissertation assesses characteristics, packaging, availability, and retailer marketing practices for these products.
Methods
A content analysis of 140 delta-8 THC product packages from the International Cannabis Policy Study was conducted to explore product characteristics, cannabinoid content labels, health warnings, and marketing appeals (Aim 1). Availability of products containing cannabinoids other than delta-9 THC and CBD was determined for 39 online CBD/hemp retailers in Maryland, and cannabinoid content information for 175 products from these retailers was reviewed (Aim 2). A content analysis of 20 CBD/hemp retailer websites in Maryland was conducted to assess age verification, health claims, warnings, promotions, and engagement strategies (Aim 3).
Results
Most ingestible products contained more than 10mg of intoxicating cannabinoids per serving. Cannabinoid content labels for packages and product listings varied substantially, with many lacking sufficient information to calculate cannabinoid concentration. Lab results often differed from information in product listings. Explicit and implicit health claims were present on most retailer websites, with 50 health conditions mentioned across websites. Warnings on packaging and retailer websites were often small, inconspicuously placed (e.g., on the back or bottom), or absent altogether. Packages and retailer websites frequently included language describing products as “hemp,” “legal,” or “natural.” Half of retailer websites lacked age verification.
Discussion
The range of cannabinoids found in this study highlights the need for regulations requiring standardized cannabinoid content labels that communicate dose clearly and consistently. The presence of unsubstantiated health claims—both explicit and implicit—highlight the need for regulations around what the industry and retailers can communicate to consumers about the potential effects of cannabis products. Large, concise, rotating health warnings are needed at the top of product packages and retailer webpages to inform consumers about potential risks of cannabis, and effective age verification is needed to curb youth access
STUDYING SPIN LIQUID CANDIDATES WITH TRIANGULAR LATTICE
No full textUnderstanding quantum entanglement as manifested in magnetic materials
is necessary to take advantage of quantum systems to overcome the classical computational limits. In condensed matter physics, the quantum spin liquid represents a new state of matter that is defined by entanglement rather than by
symmetry breaking. However, there is not yet a widely agreed upon materials
realization of a quantum spin liquid.
In this dissertation, we focus on triangular lattice antiferromagnetic (TLAFs), which have been studied because of their potential to support quantum spin liquid. We examine three different compounds, NaBaYb(BO3)2, NaYbSe2, and K2Co(SeO3)2. All have triangular lattices of magnetic ions with antiferromagnetic interactions. The dissertation presents detailed magnetic studies of these compounds using X-ray diffraction, neutron diffraction, DC susceptibility, heat capacity measurements, and other complementary techniques. A discussion of the results, as well as potential future experiments, are included
EFFICIENT PARAMETERIZATION OF SURFACE DIFFUSION KINETIC MONTE CARLO SIMULATIONS TO STUDY THE MORPHOLOGICAL EVOLUTION OF BODY-CENTERED CUBIC NANOPARTICLES
Full text linkMesoscale diffusion kinetics bridge atomic-scale interactions with macroscopic material morphology and properties, enabling the rational design and optimization of materials for diverse applications. Kinetic Monte Carlo (KMC) simulations are often preferred for these studies because they can model longer time scales, handle complex systems, and provide detailed insights into kinetic processes. This work explores the use of an extended broken bond model based on two shells of atoms to predict activation barriers of diffusion in BCC materials While effective for general insights, the broken bond model fails to capture specific material intricacies. Precise parameterization of rates for specific materials remains a challenge due to the associated computational expense.
This thesis reports a novel multiscale parameterization method for KMC that balances accuracy with computational efficiency. The process starts with a broken bond model KMC simulation to generate relevant transitions, followed by nudged elastic band (NEB) calculations in Quantum ATK to define activation barriers. These calculated values are then used to develop a three-tiered method for predicting activation barriers based on the local atomic configuration. The first tier utilizes a heuristic for predicting the activation energy involving calculating the difference between the sums of regression coefficients and the number of neighbors in the initial and final configurations across the three neighbor shells. An additional configuration-dependent constant then adjusts this difference, which accounts for the specific local atomic environment. If the configuration-dependent constant cannot be properly assigned, the model checks a fallback heuristic, which utilizes the same continuous variables but changes the configuration-dependent constant. In this case the constant is based on the hierarchical clustering of coordination numbers in the first three neighbor shells about the diffusing atom in the initial and final positions. Finally, if neither constant from the primary or secondary model can be defined for a transition we apply the model omit it from the heuristic and utilize the continuous regression of the number of neighbors in the nearest three shells in the initial and final position. This hierarchical multiscale parameterization method was implemented in KMC simulations of surface diffusion-driven morphological equilibration of body-centered cubic β-titanium nanoparticles
Au-delà des corps : immunités féminines en littérature française contemporaine
No full text“Au-delà des corps : immunités féminines en littérature française contemporaine,” examines the articulations and mutations of “immunity” in contemporary French literature as a conceptual site where discourses of embodiment, sociality, and memory merge within the artifice of fiction. While previous studies of immunity in French literature tend to derive from an archive restricted to medical novels or pathographies from the 17th to the mid-20th centuries, focusing almost exclusively on masculine characters, my project places contemporary narrativizations of mental and physical illnesses since the 1980s within a critical framework based not only in the medical humanities, but in the fields of narratology, semiology, and feminist literary theory. I argue that the way immunity shapes how health and sickness become legible in literature relates to how authors map femininity and senses of self onto a spectrum of normalcy and alterity. To substantiate this claim, I propose four distinct sub-concepts that structure discourses of immunity: healing, defense, transmission, and enhancement. In a time of mass medicalization, novel experiments in public health policy, and increased interest in bodily augmentation, contemporary fiction serves as an experimental place wherein discursive representations of abstract medical concepts can be shaped and newly deployed.
The dissertation examines four novels written by women that foreground personal and social instantiations of “immunity,” informing their structure as well as the logic of their symbolism. I open by examining the notion of healing as repair (réparation) in Maylis de Kerangal’s Réparer les vivants, asking to what extent an individual’s physical restoration via organ transplant can strengthen an entire community, going even as far as to “mend” the deceased organ donor. I then analyze Julia Deck’s exploration of immunity as a creative act across three novels wherein self-defense continuously actualizes the self, thus creating, isolating, and, ironically, exhausting the protagonists’ sense of being. Next, I turn to the interplay between medical and epistemological transmission throughout Marie Redonnet’s Les Héritières trilogy where contagion regains its etymological roots as “contact” and becomes itself a medium of communication and, ultimately, embodied knowledge. I end with a focused examination of Marie Darrieussecq’s Notre vie dans les forêts and how it traffics in transhumanist discourses, presenting a never-ending cycle of attempts to go beyond biological limitations of the human that only diminish the initial subject and impose further deficiencies. Darrieussecq’s representation of human enhancement questions an individual’s corporeal, linguistic, and social limits alike, opening up the fundamental question of human vulnerability which I broach in conclusion
DEVELOPMENT OF NOVEL COMPUTATIONAL METHODS FOR CHARACTERIZING PROTEIN ALLOSTERY
No full textMany proteins cannot be adequately understood and drugged without considering allostery. Yet allostery is often difficult to study. Proteins of interest to me, such as the SARS-CoV-2 main protease, have remained poorly understood despite extensive research. To enable faster and more powerful analyses of allostery, I developed new computational tools. I then applied them to proteins of interest to generate novel findings about those protein’s allosteric behavior. This thesis addresses two key questions related to allostery. The first is how to locate potential allosteric sites on proteins. The second is how to quantify allostery through networks of correlated dihedral angles. These questions are not merely technical challenges: they have implications for our ability to make generalized statements about allostery, as well as for the proteins that the methods are applied to.
This thesis contains several chapters, each with a different focus. The first chapter contains background on protein research in general, and on allostery specifically. The second chapter discusses the TACTICS software for finding binding sites on proteins. This chapter describes the software methodology, as well as applications to several important proteins. The third chapter discusses the EVADE software for characterizing allosteric communication. EVADE is applied to SARS-CoV-2 main protease to propose a mechanism for a known allosteric inhibitor. The fourth chapter discusses future directions. The appendix describes investigation of the free energy landscape of AMPA receptor desensitization. Together, the work presented here advances our understanding of allostery as both a general phenomenon and a regulator of the proteins discussed here
Exploring maternal immunization decision making, demand, and readiness in advance of future vaccines in pregnancy
Full text linkVaccination in pregnancy presents a critical opportunity to address infectious diseases that are especially severe for pregnant women and newborns. The landscape of vaccines designed specifically or permissively recommended for this use is expanding, yet pregnant and lactating people remain excluded from vaccine trials and subsequent policies in many lower-resourced settings. Using a mixed methods approach, this dissertation explores factors influencing decision-making for vaccines in pregnancy in low- and middle-income countries settings, with an eye toward future vaccination strategies.
The first paper characterizes archetypes among 400 pregnant and lactating women in Kenya, using latent class analysis. Considering constructs relevant to acceptance of a maternal respiratory syncytial virus vaccine, two classes emerged. “Questioners” were more likely to express distrust in vaccine benefits than “acceptor” class members. In a simple logistic regression, care-seeking at a private facility was associated with “questioner” class membership.
The second paper examines health care provider perspectives on upcoming maternal group B streptococcus (GBS) vaccines using 100 survey responses and 12 in-depth interviews. This convergent mixed methods analysis described knowledge, attitudes, beliefs, influences, and information sources and explored associations with provider vaccine hesitancy and willingness to recommend a Ministry of Health-approved maternal GBS vaccine. Provider awareness and knowledge of GBS varies, but providers anticipate demand challenges due to low community awareness. Most would recommend a maternal GBS vaccine once approved.
The third paper explores national and global perspectives on inclusion of pregnant women in clinical trials and policies. Country decision makers, researchers, and partners seek trial data demonstrating vaccine safety in pregnancy, and the absence of such trials in low- and middle-income settings is a significant barrier. Insufficient epidemiologic data of disease burden and maternal and neonatal outcomes limits countries’ ability to weigh vaccine policies. Pregnancy and antenatal visits are an opportunity to reach women with vaccines while they interact with the health system.
Together, these findings can inform tailored strategies to generate demand, improve delivery platforms, and support uptake in Kenya and other low- and middle-income countries, and to shape more inclusive research and policy agendas that center the needs of pregnant people
Probabilistic and Neural Models for Estimating Coupling Dynamics of Circadian Gene Regulatory Networks
No full textCircadian oscillators form a crucial component of biological systems. In particular, a \textbf{circadian molecular clock} is a set of genes that help regulate the behavior of an organism in a 24-hour cycle (hence the name circadian). These genes, which we refer to as clock core genes, are transcribed (DNA converted to mRNA) and translated (mRNA to protein) in a feedback-loop that is regulated by these constituent genes and external factors (called \textit{zeitgebers}). Transcription and translation cause the gene expression levels to oscillate cyclically, causing rhythmic, periodic patterns. Hence, the oscillation of several genes in a cell can be treated as a dynamical system, with the expression level of each gene being represented as a state in time, thus allowing us to leverage the rich mathematical literature that has evolved from dynamical systems, even employing signature methods, probabilistic models, and other methods to understand how these gene time series, as well as the time series of external factors that influence them, interact.
As such, in this thesis, we will unravel an essential problem in chronobiology - the \textbf{coupling} of various circadian oscillators in a biological system. We will formulate this as a dynamical system and explore methods to reproduce the dynamics using actual gene regulatory network data and infer the coupling. We will compare the efficacy of each method and discuss our computational simulations and results as a basis of better understanding how genes interact, using probabilistic machine learning, physics-informed deep learning and signature methods, as well as data-driven dynamic mode decomposition methods, as novel applications to gene regulatory networks data to estimate the coupling dynamics without the need of a specified governing differential equation
Inverse Optimization Methods for Personalized Medical Decision-Making and Treatment Planning
Full text linkThis dissertation presents a comprehensive exploration of inverse optimization methods, emphasizing their theoretical advancements and practical applications in healthcare, particularly in precision nutrition and radiation therapy. Inverse optimization has emerged as a powerful framework for learning underlying decision-making processes from observed data, but challenges such as handling noisy observations, balancing expert-driven constraints, and addressing diverse problem settings remain unresolved. This work addresses these gaps by introducing novel methodologies that bridge theoretical innovation with practical utility.
The first contribution of this work is the development of the Inverse Learning framework, a novel approach that concurrently learns optimal solutions and underlying cost functions from observed decisions. This framework incorporates the Observation-Constraint Trade-off, enabling a nuanced balance between fitting observed data and adhering to expert-driven constraints. Extensions of this framework include the Goal-Integrated Inverse Learning model, which allows for prioritizing specific constraints to explore a spectrum of solutions, and the Convex Inverse Learning model, which generalizes these concepts to parametric convex optimization problems. These models provide theoretical guarantees for non-asymptotic bounds, finite-sample robustness, and convergence, leveraging techniques such as concentration inequalities to ensure reliability under noisy conditions.
The dissertation applies these methodologies to two critical healthcare domains. First, in precision nutrition, it develops a hybrid model combining clustering and inverse optimization to generate personalized dietary recommendations that account for patient preferences and adhere to expert nutritional guidelines. Using the National Health and Nutrition Examination Survey dataset, the approach demonstrates improved dietary adherence and structural improvements in patient partitioning and utility recovery compared to standalone clustering or optimization models.
Second, in radiation therapy treatment planning, our research introduces the Inverse Optimization for Radiation Therapy (IOIRT) framework, which iteratively improves clinically accepted treatment plans using inverse optimization principles. By systematically adjusting organ-at-risk dose constraints and leveraging data-driven optimization, the framework achieves substantial dose reductions while maintaining clinically acceptable target coverage. Validation on prostate cancer patient cases highlights significant improvements in dose-volume metrics and robustness to plan degradation.
Through these advancements, the dissertation establishes a unified methodology for addressing noisy and multi-observation settings in optimization. By integrating statistical learning techniques, it offers scalable and interpretable models that outperform conventional approaches in recovering cost vectors and generating feasible, optimal solutions. The practical implications are far-reaching, providing actionable insights for personalized healthcare delivery and decision support tools
Regulation of Axon Ultrastructure
Full text linkNeurons are known for their intricate cellular processes. The neuron process called the axon is exceptionally long (100-1000 mm) and ultrathin (100 nm). The cable-like morphology of the axon is essential for electrical signal conduction, or action potential conduction, throughout the brain and body. It has long been assumed that small, unmyelinated axons are tubular structures with occasional synaptic varicosities. However, our work has challenged this assumption. Using high-pressure freezing to preserve membrane morphology for electron microscopy or super-resolution imaging of live neurons, we performed ultrastructural analysis of unmyelinated axons of mouse hippocampal neurons. We discovered that axons are not simple tubes but rather exhibit a pearls-on-a-string morphology through their entire length, with the pearls being ~250 nm and the strings ~70 nm in diameter. This morphology is reminiscent of membrane tubes undergoing tension-driven instability. Consistent with this notion, axon pearling can be changed by changing extracellular osmotic pressure, by pharmacological perturbation of the cytoskeleton, or by increasing plasma membrane fluidity through cholesterol depletion. Our in silico modeling further supports that tension-driven instability drives pearled axon morphology. In silico modeling also predicts that pearled morphology greatly impacts action potential conductance. Indeed, when axon morphology is altered by cytoskeleton perturbation or cholesterol depletion, action potential velocity is altered as predicted. Surprisingly, providing an electrical stimulation known to induce synaptic plasticity alters the pearled axon morphology, demonstrating a novel form of neuronal plasticity. These data have revealed for the first time that axons are pearled not tubular, and that pearled axon morphology has an important functional role in neuronal activity and plasticity
AN EXPLORATION OF RESOURCES AVAILABLE ONLINE FOR TEACHERS WHO WORK WITH ENGLISH LANGUAGE LEARNERS
Full text linkThe National Education Association (2020) predicts that in 2025, 1 out of 4 students in the United States will be an English Language Learner. While the number of ELLs rises, teacher training for educators that work with this population has remained stagnant, leading to teachers looking online for ways to support their ELLs. This dissertation looked at the online resources available for teachers working with ELL. Using the Patient Education Material Tool (PEMAT) (Agency for Healthcare Research and Quality, n.d.), this study reviewed the understandability and actionability of the online materials. The study also looked at the quality of the resources compared to researched best practices when working with ELLs. Lastly, the study examined the language ideologies present in the materials. The findings indicate that while most resources are understandable based on PEMAT’s criteria, they often lack actionability. Many best practices for supporting ELLs are present in the materials, with scaffolding being the most emphasized strategy. Finally, the study identifies a range of language ideologies: some resources reinforce the hegemony of the English language, a few reflect a subtractive perspective—viewing ELLs’ native language as a barrier to English acquisition—while most promote either a pluralist ideology, which supports maintaining ELLs’ native language, or an additive approach, which advocates for schools to develop students’ native language alongside English actively