22689 research outputs found
Sort by
SCIENCE FROM A TO GEN Z AN EXPLORATION IN CONNECTION
No full textOur future as a species and a planet is, inevitably, determined by the youth. They are our next generation of scientists, explorers, and humanitarians. This collection of personal essays and reported pieces, interluded by thematically relevant poems, celebrates an age group that is often publicly lambasted and broadly misunderstood. The topics represented in this thesis explore science for and about the children and young adults born between 1997 and 2012, including the author herself: Generation Z. Femininity, death, family, and connection are overarching themes
SUPPORTING U.S. WHITE TEACHERS’ CROSS-CULTURAL COMMUNICATION
Get PDFU.S. public school classrooms continue to diversify as the population of multicultural students increases annually, and yet the teaching population in most public schools has remained consistently composed of White women. With the shift in the student-body population there has been a shift in the content and subject matter taught in many schools, in order to ensure that the content is reflective of the multicultural students. Thus, many public schools have adjusted what they are teaching in multicultural classrooms but have yet to adjust how they teach multicultural students. The manner in which students are communicated with by teachers in most U.S. public schools is a communication style reflective of the White dominant culture, because often the teachers are purveyors of the White dominant culture. As classrooms become increasingly diverse the communication style taught and utilized within most classrooms is still rooted in Whiteness, and thus, the communication style used for classroom management by most teachers is not reflective of the communication style used in Black and Latine homes. While most White teachers utilize a communication style that is rooted in choices and questions, students from Black and Latine homes are more accustomed to a communication style rooted in demands. This cultural-communication mismatch can lead to ineffective classroom management and misunderstanding or code-switching for the student. Using ecological systems theory, sociocultural theory, critical race theory, BlackCrit, LatineCrit and critical translingual theory, this doctoral dissertation will explore the presence of a White style of communication within multicultural classrooms. A research-informed intervention was developed in the form of a professional learning program in order to support White teachers’ identification and understanding the nuances associated with multicultural communication styles, and its presence within classrooms
Exploring Patient Perceptions of Emerging Gene Therapies for Arrhythmogenic Right Ventricular Cardiomyopathy
Get PDFNo disease-specific therapy currently exists for arrhythmogenic right ventricular cardiomyopathy (ARVC), a progressive cardiogenetic condition conferring elevated risk for sudden cardiac death. Emerging gene therapies for ARVC have the potential to fill this gap. However, little is known about how adults with ARVC, or any other inherited cardiomyopathy or arrhythmia syndrome, appraise the risks and benefits of gene therapy research and which considerations influence their decisions about clinical trial participation.
Twenty adults with clinically diagnosed and gene positive ARVC participated in semi-structured interviews that explored perceptions of gene therapy and hypothetical decision making around gene therapy clinical trial participation. Transcripts of these interviews were qualitatively coded and abductively analyzed.
Participants exhibited enthusiasm for gene therapy with varied levels of interest in research participation. Patients considered perceived disease severity, level of adaptation to disease, level and nature of personal ARVC community involvement, coping styles, and life stage as relevant to decision making about trial enrollment. Clinical severity metrics had little association with interest in participation. Potential ethical concerns included patient vulnerability and extreme trust in clinical teams collaborating on industry-led trials. These findings are relevant for the informed consent process, particularly as some community expectations were at odds with those of researchers. Trial characteristics that mattered to participants included trial stage, perceived participation burden, degree of clinician and researcher trust, and anticipated future cost of gene therapy. Insights from this study may affect trial planning and communication with participants who have inherited cardiac conditions
The Metaphysics of Politics: Nietzsche and His Interlocutors
No full textThis dissertation pursues two questions: (1) What metaphysical assumptions undergird dominant Western conceptions of politics? and (2) How can we conceive of politics differently? Regarding the first, I trace the established conceptions of politics back to Aristotle, who could be said to have founded both the field (politikê epistêmê) and its subject of study (politiká). The image of politics I find in Aristotle’s texts – of the quintessentially human (anthropos) activity of organizing collective life through reason/speech (logos) – is predicated on two metaphysical axioms that still undergird majority of notions of what politics is, or can be: logocentrism and anthropocentrism. Regarding the second question, I reconstruct Friedrich Nietzsche’s political naturalism as a powerful critique of, and plausible alternative to, said conceptions of politics. I study Nietzsche’s distinctive inflection of two concepts central to political thought: power (a driving force of politics) and body politic (the locus of politics). Further, I examine two groups of Nietzsche’s heirs in political thought. Both have inherited his project and thereby expanded conventional images of politics, but each only partially. The first, spearheaded by Michel Foucault, follows Nietzsche in resisting the logocentric model of politics, but remains too tethered to its anthropocentric scope. The second, spearheaded by Bruno Latour, follows Nietzsche in resisting the anthropocentric scope of politics, but remains too wedded to its logocentric model. Finally, taking the Covid-19 pandemic as a case study, I explore how neo-Nietzschean political analyses, more than standard approaches in contemporary political thought, offer conceptual resources to come to terms with the volatilities of the thoroughly entangled, more-than-human world we inhabit
Design, Microfabrication, and Application of Responsive DNA Polymerization Gel Devices
No full textShape-changing hydrogels responsive to biochemical stimuli hold significant potential for applications in biosensing, smart medicine, and soft robotics. Building on our groups’ previous work on high-swelling DNA hydrogel by DNA hybridization-powered polymerization, this thesis presents a new family of multi-component DNA polymerization gels using varied polymer backbones, including acrylamide-co-bis-acrylamide (Am-BIS), poly(ethylene glycol) diacrylate (PEGDA), and gelatin-methacryloyl (GelMA). These hydrogels, assembled into custom micron-scale shapes, demonstrate a broad spectrum of mechanical properties and biocompatibility.
To overcome the lack of reversible swelling in the DNA-responsive gels our groups previously developed, a biomolecular-signal-driven mechanism enabling reversible shape changes was introduced. By incorporating additional tail domains to existing DNA growth activator strands and introducing DNA shrinking activator strands, a bidirectional response with different gel backbones for multiple actuation cycles was achieved.
Research in this thesis further used this scheme to create gel automata — micro-segmented materials capable of transforming between various predefined shapes. By integrating photolithography techniques with machine-learning-assisted design methods, unique and significant shape transformations such as letters and numbers using DNA sequence instructions were achieved. Moreover, the integration of a mechanochemical gel module with the gel automata has paved the way for adaptive, hydrogel-based robotic devices that can both sense and respond to mechanical stimuli.
This thesis also examined how to fine-tune actuation responses by altering DNA and environmental parameters, demonstrating the versatility of these systems. Lastly, this thesis presents ultra-thin graphene oxide (GO) composite DNA polymerization gels that retain unique responsiveness, while enhancing their mechanical properties and shrinking kinetics, signifying potential applications in soft robotics and soft electronics.
The advancements detailed in this work highlight the improvements and vast potential of DNA polymerization gels, setting the stage for future breakthroughs in areas such as soft robotics, drug delivery, and tissue engineering
GRAPH BASED DEEP LEARNING MODELS FOR ANALYSIS OF RESTING STATE FMRI WITH APPLICATIONS IN LOCALIZATION AND DYNAMIC FUNCTIONAL CONNECTIVITY
Get PDFResting-state functional magnetic resonance imaging (rs-fMRI) is a noninvasive neuroimaging modality that quantifies the changes in blood flow and oxygenation in the brain at rest. Analyzing connectivity graphs extracted from rs-fMRI can yield insight into the functional organization of the brain.
Neurosurgical resection procedures require immense precision, as the surgeon must remove as much of the lesion while preserving maximum functionality. An incorrect incision could cause severe or even permanent cognitive deficits. Rs-fMRI has emerged as a preoperative mapping modality for eloquent cortex localization for brain tumor removal surgeries as well as epileptogenic zone removal surgeries in patients with epilepsy.
Rs-fMRI connectivity analysis usually begins with applying an existing parcellation to define regions of interest (ROI's) of spatially and temporally homogeneous areas of the brain. However, these existing parcellations do not generalize well to patients with brain tumors or epilepsy due to an atypical rs-fMRI signature.
We present a collection of deep learning methods to analyze rs-fMRI of atypical populations, namely brain tumor and focal epilepsy subjects, to perform localization of regions of interest for neurosurgery. First, we explore techniques to develop more accurate subject-specific parcellations for downstream analysis using refinement techniques. We develop a Bayesian model with a markov random field prior to refine parcellations on a subject-specific basis. We then present RefineNet, which jointly optimizes parcellation refinement and the downstream tasks.
Then, we present our models on eloquent cortex localization for tumor patients. We leverage graph neural networks to perform localization. We extend our model using both temporal and spatial attention models applied to dynamic connectivity, where our attention mechanisms capture spatiotemporal features that boost localization performance.
Next, we present our models on epileptogenic zone localization for epilepsy patients. We develop a graph convolutional network called DeepEZ which uses anatomical connectivity regularization and a biologically inspired loss function. We extend this work to the dynamic connectivity case as well, using transformer networks to capture temporal nuances. Lastly, we tackle the noisy label problem through the context of epileptogenic zone localization, and develop a framework to perform localization even when trained on a dataset with entirely noisy labels
CANCER IMMUNOMICS: DECIPHERING THE CELLULAR AND MOLECULAR LANDSCAPES OF IMMUNOTHERAPIES
No full textThe immune system defines a complex network of tissues and cell types that orchestrate responses across the body in a dynamic manner. The local and systemic interactions between immune and cancer cells contributes to disease progression. Lymphocytes activated in lymph nodes, traffic through the periphery, and impact cancer progression through their interactions with tumor cells. As a result, therapeutic response and resistance are mediated across tissues and a comprehensive understanding of immune cell dynamics requires a systems-level approach. This body of work leverages single cell and spatial -omics technologies in combination with matrix factorization and transfer learning methods to model immune cell dynamics in preclinical experimental mouse studies and clinical trials. First, we demonstrated the use of epigenetic modulation to sensitize HER2+ tumors to immune checkpoint inhibition in a murine breast cancer model. We applied single cell RNA sequencing to elucidate the effects of epigenetic modulator entinostat and aPD1, aCTLA4 immune checkpoint inhibitors in the immune tumor microenvironment. Our results lead to a model that entinostat treatment promotes tumor killing mechanisms by promoting an anti-tumor myeloid cell signature and reducing immunosuppression in MDSC cell subtypes. The combined alterations in the myeloid compartment ultimately improved T cell infiltration and cytotoxicity. Next, to assess how mechanisms of immune response studied preclinically translate clinically, we developed CyTOFpatterns, a computational framework tailored for mass cytometry, to study temporal dynamics of patient immune responses to immunotherapy strategies. CyTOFpatterns is an NMF and transfer-learning based framework that enables unsupervised protein expression pattern learning and projection across patient cohorts for the purpose of prognostic and predictive biomarker development in the context of cancer immunotherapies. Finally, we applied spatial transcriptomics to elucidate intratumoral immune dynamics and tertiary lymphoid structure formation in pancreatic cancer patients after participating in neoadjuvant immunotherapy trials. Altogether this work builds upon computational methods to leverage cancer immunomics data resolved at the single cell and spatial levels to extract meaningful biological and translational insights that can inform potential new combination therapeutic strategies for clinical trials
Statistical Methods for Spectral CT Imaging and Material Decomposition
Get PDFSpectral Computed Tomography (CT) is a medical imaging modality that involves measuring the energy-dependent x-ray transmission properties of a patient from many view angles. These measurements can be used to reconstruct medical images that show all the same information in conventional CT
as well as additional information about the material composition of each object, due to the fact that different materials have different x-ray attenuation spectra. The goal of this research is to use mathematical models, computational simulations, and physical experiments to understand, innovate, and optimize spectral CT imaging systems for biomedical applications.
There are several key tools and models that make this research possible. First, we use a probabilistic forward model of spectral CT imaging physics which includes signal and noise transfer associated with the polyenergetic x-ray source, the mass attenuation spectra for each line of response through the patient, and the sensitivity spectra of the detector. We also rely on model-based material decomposition algorithms, which aim to estimate material density images given spectral CT measurements by optimizing a statistical objective function. Finally, this research uses quantitative metrics of image quality including covariance, bias, detectability and a new metric we call separability.
In this research, we propose new spectral CT instrumentation, material decomposition algorithms, and statistical machine learning models. We believe this work has the potential to address existing clinical challenges and accelerate new biomedical applications of spectral CT
TRANSPORT STUDY OF SUPER-PARAMAGNETIC IRON OXIDE NANOPARTICLES AS DRUG DELIVERY VEHICLES
No full textIn recent years, there has been a growing interest in the use of magnetic nanoparticles(MNPs) for therapeutic applications including drug delivery, magnetic hyperthermia, magnetic resonance imaging, and more. However, the efficacy and transport kinetics of magnetic nanoparticles w.r.t. overall size, the influence of the magnetic field, and their fate as they move across different biological barriers have not been examined in detail.
To study transport across biological barriers, Round Window Membrane (RWM) presents a unique biological barrier comprising a 3-layer cellular structure (Outer epithelium with tight junctions, middle connective tissue layer, and inner epithelium with loose cellular junctions). Another biological barrier important for vascular drug delivery systems is blood vessels (BV) which are made of a different tri-layer structure- (squamous epithelium cells, smooth muscles, and outer layer of loose connective tissue).
This dissertation describes a rational approach toward the design of monodisperse super-paramagnetic iron oxide nanoparticles (SPIONs) as drug delivery vehicles that allows tuning magnetic property with core size and/or overall size using polymer coating thickness. This research illustrates the biocompatibility of SPIONs designed for cochlear drug delivery applications up to very high iron concentrations of 0.5 mg-Fe/ml in the murine organ of Corti cultures. Transport results show increased transport efficacy of NP-PEG 3000 (~144nm) vs smaller size NP-PEG 600 (~44nm) & larger size NP-PEG 6000 (~188nm) across the acellular submucosa small intestine membrane (SIS), RWM, and BV. Transport studies show significant SPION transport across cellular membranes without magnetic field application, demonstrating native cellular processes involved in drug delivery and overall increased kinetics of transport with an applied magnetic field.
Imaging studies for RWM reveal aggregation of NP-PEG 600 during transport highlighting the importance of core-to-core separation for stability of MNPs. Membrane permeability and SPION diffusivities are calculated for each tissue type and SPIONs respectively. The results obtained in this work will pave the way for the rational design of magnetic nanoparticles as drug delivery vehicles and will be a guide for the translation of this work into in vivo studies with the final goal of their clinical application for improving drug delivery systems across several biological barriers
Applications of a Thermosensitive Hydrogel: From Cancer to Skin Rejuvenation
No full textDespite decades of advancements in therapeutic delivery options, there are still many therapies that suffer from delivery issues in a breadth of disease areas. These problems include off target toxicity, low accumulation of therapeutics at diseased sites, and poor bioavailability of therapeutic molecules. In this dissertation, I will explain the development of a thermosensitive and biodegradable poly (δ-valerolactone-co-lactide)-b-poly(ethylene-glycol)-b-poly(δ-valerolactone-co-lactide) (PVLA-PEG-PVLA) hydrogel for the delivery of therapeutics for the treatment of both an aggressive phenotype of triple negative breast cancer and the treatment of senescent dermal fibroblasts. The PVLA-PEG-PVLA hydrogel is unique in its ability to gel when reaching physiological temperature without the characteristic pH drops and has tunable features that can allow us to tailor the properties to the desired application.
The first application of this hydrogel system I will present is for the treatment of E-cadherin positive triple negative breast cancer (TNBC). E-cad expression correlates with a worse overall survival rate for invasive ductal carcinoma (IDC) patients. A recent discovery has shown that MEK inhibitors can be employed to treat this particularly aggressive phenotype of TNBC; however, these inhibitors often fail in the clinic due to severe toxicity associated with systemic (oral) administration. To address this challenge, we utilized liposomes functionalized with PR_b, a biomimetic peptide that targets the 5β1 integrin overexpressed on cancer cells and co-encapsulated a MEK inhibitor (mirdametinib) and chemotherapeutic doxorubicin. By combining these functionalized liposomes with our thermosensitive PVLA-PEG-PVLA hydrogel, we were able to locally deliver the therapeutics locally to the tumor site, which significantly improved toxicity issues and lowered the amount of drug required to successfully treat TNBC tumors in multiple in vivo models.
In the second application of the hydrogel, I will show this system can also be used to deliver therapeutics for the treatment of senescent dermal fibroblasts. Senescent fibroblasts accumulate in the dermis with age and through production of inflammatory molecules contribute to dermal dysfunction. By combining a hypoglycemic drug, metformin, encapsulated in PR_b functionalized liposomes with a hypertension drug, valsartan, mixed directly in the hydrogel, we demonstrate a promising strategy to improve the function of these senescent dermal fibroblasts. Our treatment resulted in phenotypic improvements of the senescent cells as well as enhanced collagen production in two in vitro models, providing a proof-of-concept system design for the treatment of aging skin.
Finally, I will present work aimed to modify this PVLA-PEG-PVLA hydrogel to act as a scaffold to support the growth and other functions of cells to generate tissues and functioning organs. While our attempts to improve cell viability through conjugation of bioactive peptides failed to support cell growth, we learned important lessons on how the inclusion of these molecules can affect self-assembly behavior. To address these issues, we present early proof-of-concept experiments on strategies to manipulate the polymer phase behavior to allow for conjugation of these molecules