22691 research outputs found
Sort by
NEW TOOLS FOR IMAGE-INFORMED BONE TISSUE ENGINEERING
No full textCritical-sized calvarial bone defects pose complex clinical challenges due to complicated biological processes involved in bone healing. Among these, angiogenesis is essential for delivering oxygen, nutrients, and key cell types to the defect site, yet the detailed role of vascular function on bone growth and repair over the healing cycle remains insufficiently understood. Current clinical and preclinical studies mainly focus on structural imaging and lack insight into vascular functionality and the spatial-temporal effects of angiogenesis on bone regeneration at microvascular resolution. Furthermore, the successful integration of tissue-engineered constructs in calvarial defects requires effective oxygen delivery from the vasculature to sustain construct viability and promote long-term tissue repair. Recent advancements in tissue engineering, such as proangiogenic factors and oxygen-generating biomaterials, underscore the need for continuous monitoring of intravascular and tissue oxygenation to enhance our understanding of the bone healing microenvironment.
This thesis addresses these challenges by proposing (i): a novel imaging pipeline to facilitate longitudinal, in vivo characterization of calvarial bone healing, (ii): application of this imaging framework to study the impact of tissue-engineered treatments on bone regeneration in vivo, and (iii): the development of oxygen-sensing scaffolds capable of monitoring in vivo oxygenation in conjunction with the imaging platform. This pipeline will integrate high-resolution imaging techniques to provide quantitative data on vascular perfusion and tissue oxygenation throughout the bone healing process. Through detailed oxygen monitoring, the framework aims to reveal the role of oxygen gradients in cellular processes critical to bone repair.
The results of this study are expected to yield new insights into the in vivo dynamics of oxygen delivery and vascular function in bone healing, allowing quantitative analysis and improving the design of more effective treatment and tissue-engineering strategies. By combining scaffold design with real-time oxygen sensing and imaging, this research offers the potential for significant advancement in understanding and optimizing the conditions necessary for efficient calvarial bone defect repair. These findings may guide future therapeutic approaches, improving outcomes for patients with critical-sized bone defects
Optimization and Characterization of 3D-Printed Bioreactors for Advanced Cellular Systems
No full textThis thesis presents the development and characterization of 3D-printed microfluidic bioreactors for advanced cellular systems. Current brain organoid culture methods face significant limitations due to diffusion constraints, restricting structures to diameters below 500 μm before core necrosis occurs. This challenge is addressed through systematic optimization of high-resolution microfluidic platforms that support larger tissue constructs through enhanced nutrient delivery. Using projection micro-stereolithography (PμSL) and stereolithography (SLA) printing technologies, resolution capabilities and channel clearance was characterized across multiple photopolymer resins, demonstrating that BMF Bio resin with PμSL printing achieved features as small as 250 μm with error rates below 10%. This thesis also establishes protocols for post-processing, surface treatment, and sterilization, including UV/thermal curing regimens tailored to specific material properties and PDMS compatibility requirements.Three bioreactor variants were developed: a global perfusion bioreactor incorporating a 5×5 array of microtubes with varying pore sizes, an electrode bioreactor integrating a 3×3 electrode port array for electrophysiological monitoring, and a bacterial bioreactor demonstrating crossover applications to microbial systems. The innovative "jacket" sealing strategy with integrated cell access points effectively eliminated leakage while maintaining experimental accessibility. Validation studies with neural progenitor cells demonstrated successful maintenance of viable cultures for up to 8 weeks, with immunofluorescence analysis confirming both proliferation and differentiation into mature cell types. Perfusion visualization using gelatin hydrogels verified efficient nutrient distribution throughout the culture chambers This work establishes a versatile foundation for uspporting larger, more complex tissue constructs, potentially enabling advanced models for studying neural development, disease mechanisms, and cellular interactions in more physiologically relevant microenvironments
MULTIPLE OUTPUTATION AS A REGRESSION STRATEGY FOR COMPLEX MULTILEVEL DATA
No full textThis study aims to apply multiple outputation as a regression strategy to large and complex electronic health records (EHR) data. We used EHR data from 2019–2021 across three health systems. Bootstrap, multiple outputation, and Poisson mixed effect regression were used to assess the association between the use of RTPB and other factors. To assess the estimates from multiple outputation, the method was applied to a smaller and extensively analyzed dataset and compared with estimates from other more established methods. The estimated coefficients from multiple outputation are close to those from the other methods. The data after multiple outputation remain consistent in both patient and provider characteristics. Therefore, multiple outputation could be a useful option for complex structures and large datasets
INITIAL TISSUE RESPONSE TO VOLAR FIBROBLAST TRANSPLANTATION IN NON-VOLAR DERMIS
No full textThis study investigates the early tissue response to volar fibroblast transplantation into non-volar dermis using a mouse model. The central research question is to determine the immediate effects of transplanting volar fibroblasts into non-volar skin, focusing on dermal disruption, immune and inflammatory responses, and donor cell survival. The goal is to establish a baseline for optimizing future fibroblast-based skin therapies.
Methods: Primary volar and non-volar fibroblasts were isolated from transgenic mice that express green fluorescent protein (GFP) upon tamoxifen induction. These cells were cultured in vitro to assess confluency, proliferation, and average cell size over a two-week period. For in vivo analysis, 50,000 volar fibroblasts were injected into the ear dermis of recipient mice. Non-volar fibroblasts and media-only injections served as controls. Three days post-injection, tissue samples were harvested and analyzed using Hematoxylin and Eosin (H&E) staining to assess tissue structure and inflammation, and immunofluorescence staining to visualize GFP+ donor cells, DAPI-stained nuclei, and native red fluorescence from tdT+ recipient cells.
Results: In vitro, volar fibroblasts exhibited significantly higher proliferation and confluency than non-volar fibroblasts, despite being smaller in size. In vivo, H&E staining revealed dermal disruption at injection sites across all groups, including cartilage damage, tissue expansion, increased vascular density, red blood cell extravasation, and inflammatory cell infiltration. These features suggest acute physical trauma from injection rather than cell-specific remodeling. GFP+ donor fibroblasts were successfully detected in the dermis of recipient mice three days post-injection, confirming short-term cell survival.
Conclusions: The findings indicate that three days post-injection is too early to observe biologically meaningful remodeling driven by fibroblast transplantation. The observed changes primarily reflect acute mechanical injury and early immune response. However, the ability to detect viable donor cells at this early time point establishes a clear and reproducible baseline for future studies. This model provides a valuable platform for testing how variables such as cell number and pre-treatment conditions affect fibroblast survival, immune response, and long-term tissue remodeling in skin regeneration research
INVESTIGATING THERAPEUTIC STRATEGIES FOR PREVENTION OF PATHOLOGICAL RESPONSE TO GROWTH PLATE INJURY
No full textInjuries to the epiphyseal growth plate (physis) in pediatric patients often result in the formation of a bony bridge (bony bar) that disrupts normal endochondral ossification, potentially leading to limb length discrepancies and angular deformities. Current clinical interventions, such as bar resection and fat grafting, offer limited efficacy and are associated with recurrence and incomplete restoration of growth. This thesis investigates the neurobiological mechanisms underlying pathological ossification following growth plate injury, with a focus on the role of sensory neurons and nerve growth factor (NGF)–TrkA signaling.
We developed a murine drill-hole injury model targeting the distal femoral growth plate. This model reliably produces a reproducible bony bar, enabling quantitative assessments of injury outcomes using a newly established scoring system and micro–computed tomography (µCT). Utilizing a chemical-genetic TrkAF592A mouse model and the small molecule inhibitor 1NMPP1, we selectively inhibited TrkA signaling and evaluated its effects on neurovascular invasion, osteogenic differentiation, and bony bar formation.
Our results demonstrate that sensory neurons are active regulators of pathological growth plate healing. Growth plate injury induces robust ingrowth of TrkA+ sensory fibers and associated vascularization, driving ectopic osteogenesis. Inhibiting TrkA signaling significantly reduced nerve and vessel infiltration, suppressed osteoblastic activity, and preserved cartilaginous architecture at the injury site. Transcriptomic analyses further revealed that TrkA inhibition repressed pro-osteogenic pathways while maintaining chondrogenic gene expression signatures.
These findings identify NGF-TrkA signaling as a critical mediator of neurovascular-driven ossification and suggest that its targeted inhibition may serve as a promising therapeutic strategy to prevent growth plate bridging. The work not only provides mechanistic insight into nerve-skeletal interactions during pathological repair but also establishes a preclinical foundation for future therapies aimed at improving outcomes in pediatric growth plate injuries
On-demand hypoxic hydrogels to investigate angiogenesis
No full textThe lack of oxygen, or hypoxia, elicits a multitude of tissue responses depending on the severity and duration of the exposure. Chronic hypoxia, lasting for days or weeks, is common in diseases including cancer and cardiovascular disorders. On the other hand, localized cessation of blood supply can lead to acute hypoxia, which typically lasts for a few minutes to a few hours. Our understanding of the threats caused by acute (i.e., short-term) hypoxia is limited mainly due to its transient nature and the lack of appropriate in vitro models to recapitulate its effect. Endothelial cells lining the inner blood vessels are the first to encounter changes in oxygen levels, including acute hypoxia.
In this work, we developed two hydrogel platforms that initially facilitate the formation of three- dimensional vascular sprouts and microvasculature, and then expose the three-dimensional constructs to controllable, acute hypoxia. The first hydrogel system consists of a gelatin-dextran backbone, which allows the generation of three-dimensional endothelial sprouts from spheroids and then introduces short-term moderate hypoxia “on demand”. Using the gelatin-dextran platform, we showed that moderate acute hypoxia (oxygen ~5%) leads to increased endothelial cell migration and sprouting from the established sprouts. Our findings indicate that the enhanced angiogenic response is regulated by reactive oxygen species, independently of hypoxia-inducible factors. Reactive oxygen species-dependent matrix metalloproteinase activity mediates the angiogenic sprouting that is observed following acute hypoxia.
In the subsequent work, we developed a fibrillar hydrogel based on collagen and hyaluronic acid, which enables the formation of three-dimensional microvascular tissue and then introduces severe acute hypoxia (oxygen ~1%) in the tissue microenvironment. Following reoxygenation, we observed increased angiogenesis from the established microvasculature. Time-lapse live-cell imaging of the three-dimensional vessels revealed increased mitochondrial fragmentation immediately following acute hypoxic exposure. We observed increased BNIP-3-mediated mitophagy in the “on demand” hypoxic hydrogels, which was subsequently confirmed by inhibition studies. Lastly, we examined changes in the proteome landscape of the three-dimensional vessels induced by acute hypoxia. Overall, we generated novel hydrogel platforms to study how “on-demand” acute hypoxia impacts angiogenesis, with broad applicability to developing novel sensing technologies
INVESTIGATING THE ESTABLISHMENT AND INHIBITION OF IMMUNOLOGICAL MEMORY FORMATION BY CRISPR-CAS SYSTEMS
No full textDespite the ubiquity of CRISPR-Cas technology for use in contemporary biomedical research, relatively little is known about CRISPR-Cas systems in natural models of infection by bacterial viruses, known as (bacterio)phages. In prokaryotes, CRISPR-Cas systems provide adaptive immunity by memorizing foreign nucleic acids as DNA-encoded memories called “spacers” and storing them in genomic repositories known as CRISPR arrays. In conjunction with other CRISPR-Cas effector proteins, non-coding crRNAs transcribed from spacers in the CRISPR array enable sequence-specific defense against re-encountered pathogens. Given the requirements for memory formation, we sought to explore how CRISPR systems could defend against a rapid phage infection quickly enough to memorize the invader and clear the infection.
This thesis work first describes how CRISPR-Cas systems exploit the programmed viral dormancy of temperate phages, a process known as lysogeny, to carry out CRISPR immunity. During a naïve phage infection, CRISPR-Cas immunizations are significantly enhanced against phages which choose to enter lysogeny. Further, in a population of lysogens, spacers are constantly acquired from superinfecting phages produced during spontaneous prophage induction of neighboring lysogens. Acquired spacers targeting a resident prophage cause Cas9-induced curing of the prophage, generating a population of cells heterogeneous in both spacer and prophage content. Altogether, we propose a model where the lysogenic lifecycle of temperate phages allows CRISPR-Cas immunization to occur over larger temporal windows, affording time for all steps of CRISPR-Cas immunity to take place.
Next, we analyzed how phages can overcome CRISPR immunity using anti-CRISPR (Acr) proteins. I describe a bioinformatic screen for Acrs and the functional validation and characterization of putative hits. Preliminary data on several of the putative Acrs suggest novel mechanisms for CRISPR-Cas9 inhibition.
Finally, I describe the beginnings of a project analyzing the regulation of a CRISPR-Cas system by another CRISPR-Cas locus in the same bacterium. I describe the preliminary experiments justifying further study, as well as efforts to establish assays for probing the CRISPR-based interactions in the native bacterial species
Building Locality-Aware and I/O-Efficient AI Systems
Full text linkArtificial Intelligence (AI) is becoming ubiquitous. AI models are deployed in a variety of settings and scales from large cloud and HPC clusters to the edge. Model sizes have witnessed exponential growth over the past decade. In fact, the largest models contain trillions of parameters and occupy terabytes of memory. This is driven by the need to produce more accurate models since scaling laws suggest that larger models are more accurate . However, on-device memory capacities remain limited and their growth lags behind. Device memory at the top of the memory hierarchy cannot fit large models. This is true for both training and inference. Thus, data must be moved through different memory tiers in order to perform computations on them. Unfortunately, data movement comes at a cost and is often roughly an order of magnitude slower than computation.
This dissertation investigates techniques to reduce data movement through the memory hierarchy (I/O) bottlenecks in AI workloads, borrowing from techniques used in classical computer systems design. The overarching theme of techniques presented is data locality. Data access in computer systems has been shown to exhibit both temporal and spatial locality. This property can be used to keep data that is likely to be accessed again cached in fast tiers of memory thus reducing data movement. We focus our efforts on two broad classes of problems 1) inference latency reduction in tree ensemble models and 2) efficient storage of tensors in deep learning models. Both cases involve I/O bound problems providing us with opportunities to optimize them. Optimizing I/O in both use cases enables us to achieve significant end-to-end speedups.
The first part of the dissertation presents techniques to reduce inference latency in tree ensemble models. Trees are inherently not cache friendly and their traversal incurs random I/Os. To alleviate this issue, this dissertation develops and presents two novel systems - BLOCKSET and T-Rex. BLOCKSET introduces methods to serialize and deserialize tree ensembles that optimize inference latency when models are not loaded into memory. It introduces the concept of selective access in which only the parts of the model needed for inference are deserialized and loaded on-demand into memory. BLOCKSET rearranges the nodes in a block-aligned storage format to reduce the number of I/Os for inference. T-Rex trades many random I/Os for few sequential I/O by remapping a forest of trees into a single spatial index.
The second part of the dissertation investigates efficient deep learning model storage and retrieval in large computing clusters. We present our system DStore - a lightweight, distributed, RDMA-enabled learning model repository that enables fine-grained tensor access and the ability to reuse tensors in a model in a distributed setting. Furthermore, we extend DStore to equip it with the ability to handle provenance and ancestry queries. The new system is called Evostore. We evaluate Evostore's ability to perform continuous transfer learning at scale in an end-to-end Neural Architecture Search (NAS) workload. We then extend Evostore to consider efficient usage of multiple and hierarchical storage tiers. We explore the use of access patterns in caching decisions as well as local client caching in this setting. We thus present PTStore which demonstrates that these applying techniques lead to significant speedups for multiple workflows such as NAS and prefix caching for LLM inference
Language, Equity, and Assessment in Gifted and Talented Education
No full textGifted and talented education occupies a highly contested space in research, policy, and practice because of how the field historically has been employed to withhold opportunities and perpetuate structures of exclusion and inequity. This dissertation probes the influence that language bias may have on gifted and talented identification, and it forms a narrative about how language bias is understood and experienced in the field. The dissertation opens with a literature review on equity in education and continues with three studies examining one form of inequity, language bias, in gifted and talented education. The studies are a systematic review of the gifted and talented literature, a differentiated qualitative replication interview study with educators of the gifted and talented, and an original qualitative cognitive interview study with gifted and talented students. The findings from the studies are summarized and connected in the concluding chapter.
This work revealed that language bias is insufficiently studied and inconsistently defined even though the influences of language are keenly felt by both educators and students. Moreover, the findings indicate that the field needs to unravel how it has intertwined language and culture and the assumptions related to language, culture, and bias that have underpinned previous research in the gifted education landscape. The findings also suggest that creativity and context can be in competition in testing environments, and patterns that are correlated with divergent thinking and access to relevant learning opportunities may underlie incorrect responses.
My findings add texture to the conflicted findings about the appropriateness of using such a measure for GT identification, and they challenge the commonly-held perception that nonverbal measurements are language-free measures that serve students from diverse language backgrounds equally well. Specifically, the findings counter the belief that removing written text from a test is equivalent to removing the presence of language from the testing experience. I argue that consensus around a definition for language bias is the first critical step in advancing the field, and I conclude this dissertation by offering a possible definition for the term. I also argue that the field will benefit from increasing innovation in how we conduct research, and I propose integrating more open science practices (e.g., replication) in education research, especially qualitative research, to expand our knowledge and inform best-practices for the field
Sporting a Baltimore Chinese Identity in 20th-Century Baltimore
Full text linkIn the mid-twentieth century, Baltimore’s Chinese community was no stranger to sports. Through playing in, spectating, and organizing matches and tournaments, the Chinese community engaged with sports on their own terms. Participating in the wider Baltimore sports scene provided crucial points of contact between the Chinese community and other segments of Baltimore society. A baseball match was utilized as a fundraising opportunity during the Second Sino-Japanese War, and talented Chinese individuals such as Thomas and James Goon were able not only to enter white spaces but win recognition and respect from white Baltimoreans due to their athletic prowess.
Sports played an even more crucial role in the Baltimore Chinese community by serving as a social space facilitating interactions among peers through Baltimore’s Chinese basketball and softball teams. Through practice and participation in regional tournaments, co-ethnic sports offered invaluable moments for social interaction between those of not only similar heritage and experience but also age. I argue that it was such spaces of interaction that were critical in the production and maintenance of a “Baltimore Chinese” identity among Baltimore’s Chinese youth.
Working on the history of sports in Baltimore’s Chinese community presents challenges associated with studying a small, transient population that left relatively few written records. To remedy the inadequacies of available sources, I relied on oral histories from individuals who had grown up in Baltimore in the mid-twentieth century.
Through my analysis, I aim to demonstrate that, rather than merely a tool of assimilation, sports was an avenue for Chinese youth to exercise agency by negotiating a communal identity. Furthermore, even though the vast majority of Chinese youth experienced constraints on their time, energy, opportunities, and social circles as a result of racial discrimination enacted against them and their parents, many had a vibrant social life that defied their socio-economic circumstances