Mason Journals (George Mason Univ.)
Not a member yet
3256 research outputs found
Sort by
Characterization of Extracellular Vesicles from LUHMES Cells: A Baseline for Understanding Mitophagy-Related Communication in Dopaminergic Neurons
No full textParkinson’s disease (PD) is a progressive neurodegenerative disorder increasingly understood as a mitochondrial disease. Mitophagy, the selective clearance of damaged mitochondria, is disrupted in PD, leading to mitochondrial accumulation and neuronal death. To study mitochondrial stress responses in neurons, researchers use stem cell models like LUHMES cells, a human dopaminergic neuronal line ideal for PD modeling. We hypothesize that neurons under mitochondrial stress may activate secretory mitophagy, exporting mitochondrial components via extracellular vesicles (EVs). While EVs have gained attention in neurodegeneration, EVs from healthy dopaminergic neuronal cells remain poorly characterized. Understanding how healthy neurons release mitochondrial cargo could reveal new strategies to counteract mitochondrial dysfunction in PD and related diseases. LUHMES cells were cultured using a novel protocol, and EVs were isolated via ultracentrifugation. These were characterized using western blotting for classical EV markers and mitochondrial proteins, nanoparticle tracking analysis, and immunoprecipitation for EV subtypes. Initial challenges included maintaining LUHMES viability during proliferation, prompting optimization. Western blotting suggests that healthy neurons secrete mitochondrial proteins via EVs, supporting the hypothesis of secretory mitophagy under normal conditions. This work establishes a baseline for future comparisons with PD models, and future studies will explore whether inducing secretory mitophagy could offer a therapeutic approach. Limitations include EV yield variability and weak protein signals, which are currently being optimized
Optimizing Mechanical Exfoliation of Tungsten Diselenide (WSe2) for Large-Area 2D Flakes
No full textTungsten Diselenide (WSe2), a transition metal dichalcogenide (TMD) 2D material, is a rising topic among scientists for its ability to exhibit piezoelectricity that can be incorporated into energy harvesting technology. To maximize the potential of WSe2 for energy harvesting, the material must be a couple of atoms thick—ideally a monolayer. These flakes can be achieved using cost-effective mechanical exfoliation using a tape to peel off the atomic layers from the bulk WSe2. However, no standardized mechanical exfoliation method reliably produces consistent, large-sized 2D flakes, limiting the scalability of energy-harvesting devices. In this work, WSe2 2D flakes were exfoliated using different tapes and peeling angles on SiO2 substrates of varying cleanliness levels. The optical microscope analysis revealed that polyimide tape, when peeling at 90 degrees, produced the 2D flakes as big as 100 µm. Additionally, experiments with both pre-cleaned (by acetone and isopropyl alcohol) and uncleaned substrates showed that exfoliating on a clean substrate tends to produce flake sizes approximately twice as large as those obtained on an uncleaned substrate. This work highlights the effectiveness and low cost of this method and demonstrates its potential to produce scalable WSe2 flakes for future industrial energy-harvesting applications
Tungsten Thin Film Electron-Beam Evaporation for Electronic Device Fabrication Applications
No full textTungsten (W) is a heavy metal valued for its high melting point, electrical conductivity, and mechanical strength that makes it an ideal candidate for electronic device applications. While various physical vapor deposition (PVD) techniques like sputtering have enabled deposition of W thin films, electron-beam (e-beam) evaporation features directional deposition of metallic thin films, thus allowing for lift-off patterning which is a crucial process in fabricating micro- or nanoscale devices. However, Tungsten’s low vapor pressure makes it difficult to evaporate in a traditional process condition because it requires a very high-power output. This project sought to develop a process recipe for W e-beam evaporation by focusing on tungsten’s sublimation point (around 1,700C) and using a graphite liner that helps with thermal isolation. The results show that tungsten of up to 30 nm in thickness was successfully deposited on a silica-coated silicon wafer at the 57% power level. To assess the possibility of device fabrication that involves W patterning by lift-off such as phase change memory, a CAD layout with features as small as 10 microns was created and tested for photolithography followed by W evaporation and lift-off. This study will serve as an experimental foundation for designing and building nanoelectronic devices that are made of hard-to-evaporate metals
Evaluating Intel TDX for Secure, Scalable Verification of Academic Replication Packages
No full textReproducibility is fundamental to scientific progress, yet verifying replication packages often relies on ad hoc, resource‑intensive workflows. Standard editorial and conference processes lack a unified, confidential framework for running proprietary or open‑source code at scale. We propose leveraging Intel Trust Domain Extensions (TDX) within Google Cloud Confidential VMs to securely evaluate replication packages from all Management Science articles published in 2024. Building on recent advances in confidential computing and containerization, our approach automates the deployment of R, SAS, and other environments inside TDX‑protected enclaves. In a pilot study of 352 packages, we document preliminary performance benchmarks, estimate per‐package cloud costs, and identify key technical challenges—such as handling non‑open‑source binaries via Docker wrappers—and prescribe best practices for library management, data access, and attestation. Our methodology also incorporates a collaborative, peer‑reviewed workflow to flag failures and share fixes in real time. Preliminary data suggests this framework can substantially reduce manual overhead while improving runtime transparency, with evaluated packages demonstrating mean computational costs of 0.22-0.27-$7.69 per replication), while Google Cloud demonstrates superior scalability characteristics and more robust handling of complex multi-language dependency chains. These results indicate significant potential for journals to adopt this approach over current isolated sandbox methods, though comprehensive validation across the full dataset is still necessary
Evaluation of Intel’s Trust Domain Extensions (TDX) for Replicating and Validating Research in Management Science
No full textReproducibility of scientific research is vital for maintaining the integrity of academic disciplines. Currently, journals and conference organizers rely on varying standards and platforms to evaluate and ensure the reproducibility of research. Additionally, current methods using unprotected environments risk security and integrity issues. This project assesses Intel’s Trust Domain Extensions (TDX) framework, a CPU-level technology for implementing a trusted execution environment, as a potential standardized solution for replicating and validating research findings in the business, economics, and management fields. Specifically, we perform a comprehensive evaluation of replication packages from articles published in Management Science in 2024 using TDX, leveraging major providers such as Google Cloud and Microsoft Azure. Our analysis focuses on four core aspects: 1) performance, measured in terms of replication success rates and accuracy compared to the paper’s results 2) costs, including computational resource usage as a result of runtime; 3) identified best practices in replication methods and software environments; and 4) technical challenges encountered during replication, including data availability and software dependencies. Our team’s results yielded high success rates, with few failures stemming from unavailable datasets and software license restrictions. We also saw relatively low credit usage per paper replicated ($1.33), and, on average, around a 2-hour runtime, proving that TDX may be a dependable solution for replication purposes
Data-Driven Ranking of African Food System Transformation Using Dimensionality Reduction
No full textAssessing the transformation of food systems across African countries is important for advancing sustainability and economic growth. However, measuring progress remains challenging because of the complexity of indicators and lack of standardized, objective metrics. Existing composite indices often rely on arbitrary weightings that can introduce bias and obscure genuine performance patterns. To address this problem, we present a data-driven ranking approach based on dimensionality reduction and clustering techniques. We construct a country–indicator matrix from binary values, representing whether countries meet specific food system transformation CAADP benchmarks. Next, Principal Component Analysis (PCA) was applied to reduce the dataset’s dimensionality, identifying principal components that reflect countries’ broad progress across domains like agricultural innovation, market integration, nutrition outcomes, environmental sustainability, and governance capacity. Multiple Correspondence Analysis (MCA), which is well-suited to binary data, is also used to help reveal patterns of similarity between countries based on their shared transformation characteristics, offering additional insight that may not be fully captured by PCA. Furthermore, unsupervised learning techniques such as k-means and hierarchical clustering are applied to identify performance groups with similar transformation profiles. This facilitates the classification of countries into groups such as leaders, emerging performers, and laggards, based on underlying structural similarities rather than pre-defined thresholds. Overall, the approach offers a replicable and objective framework for monitoring food system transformation in alignment with the goals of CAADP and Agenda 2063
Exploring Urinary EV Gene Expression for Early Detection of Bladder Cancer
No full textBladder cancer is one of the most prevalent malignancies affecting the urinary system, where early detection plays a crucial role in improving clinical outcomes. Urinary extracellular vesicles (EVs), which contain molecular cargo such as mRNA, offer a promising non-invasive source of biomarkers in early diagnosis. In this study, we analyzed a gene expression dataset derived from urinary EVs to identify potential diagnostic markers for bladder cancer. The analysis focused on four genes: LASS2, GALNT, ARHGEF39, and FOXO3. Notably, LASS2 and GALNT1 are expressed in cancer patient EVs, whereas ARHGEF39 and FOXO3 were expressed only in non-cancer controls. All four genes have previously been implicated in tumor-related pathways. After filtering out low-variability probes, we performed bioinformatic analyses and visualized gene expression patterns using dimensionality reduction and clustering tools. The results revealed a clear separation between cancer-associated and non-cancer-associated gene groups, suggesting that LASS2 and GALNT1 may serve as reliable biomarkers for early-stage bladder cancer. This study highlights the potential of urinary EV-derived mRNA profiling as a powerful, non-invasive approach for cancer detection and patient stratification
Evaluating Efficiency for Integrating Large Language, Small Language and Computer Vision Models into a Data Pipeline for an Autonomous Mobile Rover
No full textAn Application Programming Interface (API) is a script that connects various programs to communicate with one another through the use of endpoints. A few common APIs include: FastAPI, designed for developing high-performance, low-latency APIs; Flask, a lightweight and flexible microframework; and Bottle, an even smaller microframework without external dependencies. However, it is crucial to understand how these APIs perform in the context of LLMs, SLMs, and CV models for developing high-efficiency APIs, crucial for mobile autonomous rovers as tested in this project. To create the data pipeline, a GUI was developed using Streamlit, serving as a base for the APIs, which implemented the POST method when called by the app, thereby returning either waypoints for the rover or obstacles from the rover’s camera (APIs were used to bridge an SLM, LLM, and two CV models to the app). We use Apache’s Jmeter to stress test the APIs, where 100 virtual users were introduced over 20 seconds to perform the test 10 times. For the SLM we find response times of 2228.5 ms for BottleAPI, 145591.9 ms for FastAPI, and 156639.9 ms for FlaskAPI for the SLM, while for the YOLO model, we found response times of 1533.6 ms for BottleAPI, 15861.6 ms for FastAPI, and 12359.7 ms for FlaskAPI. However, BottleAPI was unable to handle many requests in both cases. It was therefore concluded that FastAPI and FlaskAPI seemed to be the most efficient for collecting waypoints via language models, and obstacles via Computer Visions Models, respectively
Influence of Expanded Field-of-View on YOLO-Based Car Detection Performance in Embedded Vision Systems
No full textReal-time object detection under embedded-vision constraints is central to autonomous navigation and smart-city sensing, yet single-camera systems suffer from limited field-of-view (FOV) and degraded performance in low-light scenes. To address this limitation, we evaluate a dual-Picamera2 rig with a combined 180° FOV against a conventional single-camera setup (90° FOV) for car detection on the COCO dataset using YOLOv7-tiny and YOLOv11-nano. We systematically vary ambient-light levels (800 lx, 250 lx, 20 lx) and measure end-to-end latency on a Raspberry Pi 5. Both configurations are trained on identical image subsets and inferred at 640 × 480 px. We explore mean average precision (mAP) vis-à-vis latency and power trade-offs, given a variety of illumination levels and model variants. A reproducible data processing pipeline is contributed to the literature that helps analysts’ decide when widening FOV yields greater benefit compared to merely upgrading to a newer YOLO variant on resource-constrained edge hardware
Characterizing Material Outgassing Risks in the NASA Landolt Space Mission for ODAR Compliance
No full textThe NASA Landolt Space Mission is a $19.5 million initiative aiming to put an artificial star in orbit to support precision ground-based flux calibration. In preparation, materials must comply with NASA’s outgassing requirements, which regulate gas release within the vacuum of space. This is especially important due to the mission’s rideshare status as gaseous material can condense on the sensors of neighboring missions in the launch vehicle. NASA requires submission of an Orbital Debris Assessment Report (ODAR) under NPR 8715.6—NASA’s Procedural Requirements for limiting orbital debris—detailing outgassing risks. To satisfy “Thermal Vacuum Stability” as defined in NASA-STD-6016C (Standard Materials and Processes Requirements), materials must demonstrate a Total Mass Loss (TML) of ≤ 1.0% and a Collected Volatile Condensable Materials (CVCM) of ≤ 0.1%. Given the direct exposure of Landolt’s Multilayer Insulation (MLI)—a series of radiation and thermal shielding materials—to the vacuum of space, we identified MLI materials as a primary concern for outgassing. We compiled a list of common MLI components from the NASA Multilayer Insulation Material Guidelines to prepare a catalogue with outgassing data on each material. We sourced TML and CVSM from the NASA Outgassing Database. We included materials failing to meet the thresholds for TML and CVCM in the ODAR as outgassing risks, which will necessitate a future vacuum bake-out. Our rigorous assessment of outgassing risks will both ensure the Landolt Mission’s compliance with NASA’s protocols and longevity of the program