Virginia Tech - Wake Forest University School of Biomedical Engineering & Sciences
VTech Works (Virginia Tech)Not a member yet
122049 research outputs found
Sort by
Journal of Future Foods
No full textThe food industry faces significant barriers to adopting automation, as most of the food service, retail, and processing operations are small businesses that lack the financial capacity to invest in conventional industrial automation systems. Food portioning is a fundamental operation across food industry sectors, yet it remains highly labor-intensive for small businesses. Existing automated portioning systems are generally designed for single-product, large-scale processing, rendering them financially prohibitive and operationally inflexible for small-scale operations with diverse product requirements. Advancements in artificial intelligence (AI) provide promising avenues for the development of cost-effective automation systems for small food businesses, offering adaptable solutions capable of handling multiple food types with flexibility and precision. This study proposes an AI-driven, low-cost food portioning framework as a proof-of-concept solution that integrates weight sensing with vision-language-action (VLA) control to enable adaptable handling of diverse food products. The system employs You-Only-Look-Once (YOLO)-based vision models to interpret digital scale readings while coordinating robotic picking mechanisms that transfer food items until the target weight is reached. Three vision-language models, namely Action Chunking with Transformers (ACT), OpenVLA with Optimized Fine-Tuning (OpenVLA-OFT), and π0, were evaluated on shrimp (30g), grapes (50g), and garlic (20g), demonstrating adaptability across diverse food types. The π0 model achieved a 100% success rate using only 30–50 demonstrations per food type and demonstrated efficient operational performance (e.g., 15.23 seconds to portion 30 g of shrimp). This framework demonstrates the potential for adaptive automation in small-scale food businesses, providing a preliminary foundation that addresses single-product automation limitations in food packaging, distribution and service operations.Accepted versio
ROPES Hub Research Brief: Strengthening STEM Pathways – Lessons Learned Supporting Low-Income S-STEM Students
No full textDoes Location Matter: Analyzing the Impact of Geographic Variation on Adoption Rates For Shelter Dogs
No full textUnwanted companion animals are a significant problem in the United States, and the numbers of dogs entering the sheltering system has been increasing recently.
The interplay of numerous organizational and animal factors, such as where a shelter is located in the country, its population density, and the type of organization as well as the size and age of the dog, how it arrived to the shelter, and its outcome, can significantly impact the length of stay for dogs in shelters. The current study utilized records from 2023 collected by Shelter Animals Count, a nonprofit organization that maintains a centralized database of United States animal sheltering data, to investigate whether the lengths of stay of dogs residing in U.S. differ based on the shelter’s geographic region or density of the population in that area to better understand how to address the time animals reside in shelters. The findings underscore a consistent trend across multiple regions: suburban shelters tend to achieve shorter stays for dogs compared to those in urban and rural areas with differences found by organization type. Generally, dogs’ lengths of stay at Northern, Midwestern and Western shelters are shorter than dogs residing in shelters in other regions across the United States, although these stays differed by the type of organization. Furthermore, we found that a dog’s size and its age affected its time in the shelter, such that older dogs have increased lengths of stay. This research provides a foundation for future study and offers an overview of the impact of geographic and animal variation on positive outcomes for dogs living in animal shelters.MAL
Investigation of Phase-Separated Polymeric Materials via X-ray Scattering Techniques
No full textThis dissertation presents a discussion of a variety of polymeric materials and their characterization via ultra-small, small, and wide-angle X-ray scattering methods (USAXS/SAXS/WAXS). Specifically, these experimental techniques are employed to investigate the phase separated morphologies that impart unique properties to each class of materials and allow for their use in advanced applications. It is crucial that the morphologies of polymeric materials are thoroughly characterized so that robust structure-property-processing relationships can be constructed. Chapter 2 of this dissertation discusses the fundamental principles of X-ray scattering, which form the basis for more complex experiments discussed throughout this dissertation. While historical, these foundational concepts of X-ray scattering are critical to highlight, due to instruments become increasingly more accessible and user-friendly. In Chapter 3, the origins of X-ray scattering in ion-containing polymers are discussed as it pertains to their unique hydrophilic/hydrophobic phase separated morphologies. These morphologies are becoming increasingly important to understand as ion-containing polymers increase in their commercial use. Chapter 4 discusses the use of USAXS experiments to understand the droplet size distributions of a filler in a polymer matrix and reveals that the liquid droplets in these composites can uniquely change size depending on their volume fraction in the composite. Chapter 5 discusses the use of SAXS and WAXS experiments to investigate the morphology of proton exchange membranes in hydrogen fuel cells. Ultimately, the results from X-ray scattering experiments provide strong evidence that commercial polymers used to make these membranes can mix at the molecular level to form robust membranes. Chapter 6 presents the application of SAXS and WAXS experiments to probe morphologies of 3D-printed polyimide materials. Using a combination of scattering, microscopy, and computational methods, the differences between two chemical variants could be understood, and the origins of void formation were uncovered. Finally, in the Chapter 7 of this dissertation, a selection of other polymeric systems were studied via X-ray scattering methods. Since most of the work presented in this dissertation was collected on a laboratory scale instrument, these experiments display the capability of benchtop instruments to characterize a range of different materials without the need to access a beamline.Doctor of PhilosophyAll plastics are polymers, and the term 'polymer' is defined as 'many parts or segments.' A single polymer is a long molecule consisting of multiple, repeating subunits known as 'monomers' which are defined as 'single units.' Many polymeric materials exhibit a phase separated morphology, meaning that certain regions of the material can become spatially separated from other parts of the material, like oil and water. This phenomenon gives many types of plastic their unique properties. In Chapter 4 of this dissertation, a soft polymer is blended with small droplets of metal, and the phase separation of the metal from the polymer provides the material with simultaneous electrical conductivity and flexibility. In Chapter 5 of this dissertation, polymer membranes are discussed for applications in clean energy technologies, and the phase separated structure of the membrane is responsible for the membrane's ability to transport protons. In Chapter 6, strong and temperature resistant materials are discussed and again, their properties in part are traced back to phase separation within the material, in this case through the presence of crystallinity. In Chapter 7, the phase separation phenomena of many materials classes are explored, ranging from aerospace plastics to fibers to biological polymers.
While there are many ways to study the phase separated structures of polymeric materials, one of the best techniques involves using X-rays. X-rays are a form of electromagnetic radiation, like sunlight, but much have a much higher energy. X-rays can interact with materials based on the internal composition and structure of the material, which many may be familiar with from medical imaging. In this case, X-rays can help produce an image of one's hand, revealing the internal structure of the body. In addition to absorbing or passing through materials, like one's hand, X-rays can also be scattered by materials, and X-ray scattering is the method that is used in this dissertation to study the internal structure of the materials on length scales smaller than the width of a human hair. Ultimately, throughout this dissertation, the power of X-ray scattering experiments is revealed as the experiments help provide information on many different phase separated polymeric materials. This improved understanding of materials at small length scales can help improve the design, processing, and performance of plastic materials in fields ranging from flexible electronics to clean energy membranes to aerospace materials
An Examination of Trends in the Rates of Low-Value Opioids Prescribed for Acute Low Back Pain in Rural vs. Non-Rural Virginia
No full textBackground: The Centers for Disease Control and Prevention (CDC) recommends against the use of prescription opioids for most types of acute pain. Despite these recommendations, some evidence suggests that opioid prescribing for acute low back pain (LBP) - among the most common acute pain complaints - persists. This study evaluated trends in low-value opioid prescribing for acute LBP among patients residing in rural versus non-rural areas of Virginia during 2019-2021 and evaluated the influence of the COVID-19 pandemic timeframe on prescribing rates.
Methods: In this retrospective cohort study, we examined insurance claims from the Virginia All-Payer Claims Database for adults continuously enrolled in Medicaid, Medicare Advantage, or commercial plans from 2019 to 2021. We used the Milliman MedInsight Health Waste Calculator to identify low-value claims and calculated annual and bi-monthly prescribing incidence rates per 1000 patients. Heterogeneous difference-in-differences models generated incidence rate ratios (IRRs) to express the difference in the rate of low-value opioids for acute LBP observed during the first two years of the COVID-19 pandemic (2020-2021) versus expected incidence based on the pre-pandemic timeframe (2019). IRRs were stratified by rurality.
Results: Among our cohort (n=853,775), 1,338,371 claims for opioids for acute LBP were identified, 73.9% of which were low-value. The annual prescribing of low-value opioids for acute LBP declined by 30.6% from 2019 (155.0 claims per 1000 patients) to 2021 (107.5 claims per 1000 patients) compared with the expected decline (model-predicted) of 18.6% during this period. During 2020-2021, low-value opioid prescribing for acute LBP was 79.6% of expected incidence (IRR: 0.80, p<.001). Low-value opioid prescribing for acute LBP was 0.74 times higher in patients residing in rural versus non-rural areas throughout 2019-2021 (IRR: 1.74, p<.001), and the difference in low-value prescribing between rural and non-rural patients did not change significantly during 2020-2021 (IRR: 1.02, p=.060).
Conclusions: Most opioids prescribed for acute LBP among this large, multi-payer Virginia cohort were low-value. The COVID-19 pandemic timeframe (2020-2021) was associated with an accelerated decline in low-value opioid prescribing for acute LBP. Persistent rural disparity in low-value opioid prescribing for acute LBP highlights the need to examine underlying drivers to reduce low-value prescribing and promote equitable, high-quality acute pain care.Master of ScienceOpioids are a class of medications used to treat diverse types of pain. National guidelines recommend against the prescribing of opioids for most types of acute (short-term) pain since they are no more effective than non-opioid treatments and are associated with substantial risks. Despite these recommendations, there is some evidence that clinicians continue to prescribe opioids for acute pain.
To better understand recent patterns of opioid prescribing for acute pain, we analyzed insurance claims for over 800,000 adults living in Virginia in 2019-2021. We specifically studied claims for opioids prescribed for acute low back pain (LBP), one of the most common types of acute pain treated in outpatient healthcare settings. We used proprietary software (Milliman MedInsight Health Waste Calculator) to categorize opioid claims as low-value (inconsistent with professional guidelines) or clinically appropriate. We assessed changes in patterns of low-value opioid prescribing for acute LBP throughout 2019-2021, assessed prescribing variation among patients living in rural and non-rural areas, and evaluated prescribing patterns in the context of the COVID-19 pandemic during 2020-2021.
Our cohort received nearly one million low-value opioid prescriptions for acute LBP during the 3-year study period. The prescribing of low-value opioids for acute LBP declined throughout 2019-2021, with the rate of decline during 2020 and 2021 greater than that observed in 2019. Rural residents received significantly more low-value opioid prescriptions for acute LBP than non-rural residents throughout 2019-2021, as the pandemic timeframe did not influence the incidence rates in a significantly different manner by rurality. Declining rates of low-value opioid prescribing for acute LBP are encouraging, but rural disparity points to systematic obstacles or entrenched prescribing practices. Future research should explore why rates of low-value opioid prescribing for acute pain vary by rurality to inform future efforts to mitigate prescription opioid-related harm
Simulating U.S. Presidents for a Friendly Chat: Applying Generative AI to Study Political History
No full textAdvancements in generative artificial intelligence have created new opportunities to develop large language model (LLM)-based simulation models. By designing distinct personas and training them with relevant information, modelers can simulate a wide range of agents, representing diverse personalities, socio-economic backgrounds, and demographics. The potential of these simulation models, often referred to as generative agents, extends beyond creating average representations of groups; they can also be tailored to simulate specific individuals, predicting their responses or opinions under various scenarios. In this study, we take on the challenge of simulating 60 U.S. presidents to demonstrate how this approach can contribute to the study of political history. We simulate 60 generative agents using an LLM (GPT o1) primed on the inaugural addresses of presidents from 1789 to 2025. We then ask each simulated president the question, “what factors influence the economy?” We validate the simulated responses with other LLMs tasked with predicting which president is most likely to have given each response. We then use a causal loop diagram generation tool called SD Bot to extract variables and relationships from the text responses and depict mental models. Finally, we quantify and visualize presidents’ relative similarities to each other as a network
Design and Synthesis of 8-Trifluoromethyl-Substituted Heterocyclic Small Molecule Mitochondrial Uncouplers for the Treatment of Metabolic Diseases
No full textDoctor of PhilosophyMetabolism occurs mainly in the mitochondria, or "the powerhouse of the cell." Here, nutrients and sugars from food are broken down and converted into energy through a process called mitochondrial respiration. Through this process, mitochondria provide over 90% of cellular energy. However, during inflammatory events, mitochondrial respiration is less efficacious and could cause or progress diseases such as obesity, type 2 diabetes, or metabolic dysfunction-associated liver disease (MAFLD). In response to the progression of these diseases, cells activate uncoupling proteins to increase the rate of mitochondrial respiration and reduce inflammation.
Thus, small molecule mitochondrial uncouplers have been developed to mimic the effects of uncoupling proteins and reduce inflammation.
Our lab has focused on designing small molecule mitochondrial uncouplers to treat metabolic diseases. We take inspiration from the efficacious uncoupler, BAM15. BAM15 has proven to be effective against obesity, type 2 diabetes, liver disease, and several other conditions in animal models. However, BAM15 has not been in clinical trials due to its poor druglike properties, such as circulating drug concentration and half-life. Herein, we disclose four mitochondrial uncoupler designs that improve the druglike properties of BAM15, while maintaining its efficacy. We report the discovery of N-substituted 8-trifluoromethyl-9H-purin-6- amines where lead compound SHK1112218 was identified as a highly potent mitochondrial uncoupler. Additionally, in an obesity reversal study in mice, treatment with lead compound SHK589 was observed to decrease body fat and maintain muscle mass. Together, these findings encourage the further exploration of mitochondrial uncouplers for the treatment of metabolic diseases
Fluid and Pressure Dynamics in Natural and Engineered Coastal Aquifer Systems
No full textCoastal aquifers are increasingly impacted by groundwater depletion, seawater intrusion, and land subsidence driven by long-term pumping. This dissertation uses 3D numerical modeling to evaluate how variable-density flow and geological heterogeneity influence pressure response, intrusion geometry, and deformation in stressed coastal systems. Three aquifer domains are examined: homogeneous aquifers, confined aquifers with continuous clay layers, and heterogeneous aquifers containing discontinuous clay layers (DCLs). Results show that geology strongly governs intrusion patterns. Homogeneous systems produce broad inland intrusion, continuous clays enhance vertical upconing, and DCLs create irregular and asymmetric intrusion zones. Because seawater is denser and less viscous than freshwater, saltwater cases exhibit larger and more persistent drawdowns, increasing modeled subsidence by 0.2 to 0.5 m after 100 years of pumping. The dissertation also evaluates Managed Aquifer Recharge (MAR) through analysis of the Sustainable Water Initiative for Tomorrow (SWIFT) pilot program in the Virginia Coastal Plain. The Potomac aquifer overlies crystalline basement rock, raising concern about downward pressure propagation in the context of injection-induced seismicity. Ensemble simulations reproducing the 2018 to 2022 pilot injections show that injection rates near 2 million gallons per day may generate pressure increases of approximately 40 kPa in the upper 200 m of the basement, although this response remains localized to within 2 km of the injector. Finally, models incorporating newly identified heterogeneity demonstrate that 20 m thick clay interbeds and laterally extensive DCLs significantly reduce pressure transmission to the basement, improving the stability and safety of MAR operations.Doctor of PhilosophyCoastal groundwater supplies are under growing stress due to long-term pumping, which can lower water levels, draw seawater inland, and even cause the land surface to sink. This research uses advanced computer simulations to better understand how these processes unfold underground and how local geology influences their severity. The study examines three types of coastal aquifers: uniform sandy systems, aquifers separated by thick clay layers, and more complex systems with patchy clay layers. The results show that underground geology plays a major role in shaping how seawater moves inland. In uniform aquifers, seawater spreads gradually inland. In systems with thick clay layers, pumping tends to pull saltwater upward from below. Where clay layers are irregular or broken, saltwater intrusion becomes uneven and unpredictable. Because seawater is heavier than freshwater, it also causes greater and longer-lasting pressure declines, which can increase land subsidence over time. The research also evaluates a groundwater recharge project in Virginia that injects treated water back underground to help restore water levels. Computer models of the pilot project suggest that injection can raise pressure in deeper rock layers, but the effects remain localized near the injection wells. Importantly, newly identified clay layers in the subsurface help limit how far pressure spreads downward, improving the overall stability and safety of the recharge program. Overall, this work highlights how differences in underground geology strongly influence coastal groundwater behavior and provides guidance for managing pumping and recharge projects more safely and sustainably
Legislative Studies Quarterly
No full textThis article follows the arc of congressional competence through the development and decline of the House Appropriations Committee (HAC) Surveys & Investigations (S&I) staff, an enduring oversight unit whose investigations were unobserved by design. S&I was quietly reorganized out of existence in December 2024. Why? Using committee reports, archival documents, and interviews this article shows how Chairman Clarence Cannon's 1943 design for S&I shaped its distinctive nonpartisan, under-the-radar oversight; filling a once-powerful niche made untenable in recent decades by partisan conflict. Time-series analysis of S&I study titles from 1975 to 2024 shows that the interplay of subcommittee chair-ranking disagreement and divided government constrained bipartisan S&I oversight. The conclusion situates S&I's history in research on congressional oversight, grounding abstract models, tracing institutional change, and revealing unexpected consequences.Submitted versionThis is an invited submission for a special issue of LSQ on Congressional Reports
Integrative Genomic Approaches for Plant Trait Discovery, Fungal Pathogen Identification and Taxonomy
No full textGenomics underpins crop improvement and plant health surveillance, yet its application remains constrained by incomplete detection of complex genetic variation, difficulty recovering pathogen genomes directly from diseased tissues, and inconsistent reference resources for fungal and oomycete identification. This dissertation integrates long-read sequencing, metagenomic approaches, and genome-based classification to address these limitations. First, long-read resequencing of 29 food-grade soybean and edamame genotypes enabled discovery of structural variants (SVs) associated with agronomic and seed-quality traits and established a workflow for SV candidate discovery and marker development. Experimental validation confirmed SV-phenotype relationships, including a 1,443-bp deletion between Kunitz trypsin inhibitor (KTI) genes associated with reduced expression and decreased seed KTI content. Second, long-read metagenomic sequencing was applied to vascular streak dieback, an emerging disease of woody ornamentals in the U.S., in contexts where culturing is infeasible. Sequencing of 106 samples from 34 host species across seven states identified Ceratobasidium sp. as the only pathogen consistently detected across samples and made it possible to assemble 17 high-quality genomes. Comparative phylogenomics and pangenome analyses indicated that U.S. isolates form a distinct cluster relative to Ceratobasidium theobromae and revealed gene-content differences, including candidate effectors and secondary metabolite gene clusters, which may contribute to host interaction and support improved diagnostics. Third, this dissertation introduces Myco-genomeRxiv, a web platform implementing an ANI-based Life Identification Number (LIN) system for genome-based identification and strain typing of fungi and oomycetes. Populated with 19,155 genomes from the NCBI Assembly database, the system uses genome-based classification to flag misassigned taxonomic identifiers and likely contamination and circumscribes 17,702 putative species using existing genome membership or a provisional 99% ANI threshold. Collectively, these studies integrate long-read sequencing, metagenomics, and genome-scale classification into a unified framework that expands discovery of trait-associated variation, enables genome-resolved investigation of disease from complex plant samples, and improves the stability and reproducibility of fungal and oomycete taxonomy for agricultural, clinical, and biosecurity applications.Doctor of PhilosophyImproving crops and managing plant diseases increasingly depends on rapid, accurate genomic analysis. However, many agriculturally important crop variants are missed by short-read sequencing, and numerous plant diseases are caused by fungi and oomycetes that are difficult to culture and to identify reliably due to incomplete or inconsistent reference databases. This dissertation applies long-read sequencing, direct DNA sequencing from diseased plant tissues, and genome-scale classification to address these limitations and to make genomic information more application-relevant for breeding, diagnostics, and biosecurity. In the first part, long-read whole-genome resequencing was used to characterize structural variants (large DNA changes, including insertions and deletions) in 29 soybean genotypes and to link these variants to agronomic and seed-quality traits. This work generated a high-coverage long-read dataset, discovered previously unrecognized structural variants associated with phenotypic differences, and experimentally validated specific deletions that alter gene expression and seed composition. In the second part, long-read metagenomic sequencing, direct sequencing of DNA from symptomatic plant tissues, was applied to vascular streak dieback of woody ornamentals in the United States. Sequencing 106 diseased samples from seven states enabled consistent detection of the associated Ceratobasidium sp. fungus and recovery of 17 genomes. Finally, the dissertation introduces Myco-genomeRxiv, a web platform that identifies fungi and oomycetes using whole-genome similarity. Including 19,155 genomes, the system supports rapid, standardized identification and strain-level typing, and helps detect mislabeling and contamination in public genome collections. Overall, this work demonstrates that integrating long-read sequencing with genome-based identification can enhance trait discovery for crop improvement, support genome-resolved investigation of emerging diseases directly from plant samples and provide a more robust framework for fungal and oomycete identification in research, agricultural, and biosecurity settings