Virginia Tech - Wake Forest University School of Biomedical Engineering & Sciences
VTech Works (Virginia Tech)Not a member yet
122049 research outputs found
Sort by
Modeling Arthropod Traits in a Bayesian Framework
No full textAs ectothermic organisms cannot fully regulate their own body temperature, they are dependent on external sources of heat. Because of this, temperature has a large influence on many of their physiological traits that affect fitness, such as development time or lifespan, which in turn, alters their ability to transmit disease-causing pathogens. The effect of temperature on traits is of interest as it can aid in developing effective control and mitigation strategies under changing climates. Researchers have been able to model this relationship using thermal performance curves (TPCs), or curves that quantify how an organism's performance changes as a function of temperature. However, most thermal analyses have focused on finding the ``best" TPC equation, often overlooking other key considerations. The purpose of this work is to use Bayesian methods to fit the most accurate TPCs with the least uncertainty based on data availability, as well as investigate certain overlooked pieces. Specifically, we conduct simulation experiments to explore the effect of the data-generating mechanism, or the distribution of the data underneath the curve. Additionally, we fit hierarchical models in two scenarios, within a genus and between genera, to explore the potential for information-sharing and generalization between species to improve curve estimation. Lastly, we extend these hierarchical models in a single-species context to incorporate relative humidity directly into the TPC through its parameters, providing an effective framework for modeling multiple environmental stressors simultaneously.Doctor of PhilosophyEctotherms, often referred to as ``cold-blooded" organisms, are organisms that cannot fully regulate their own body temperature, such as snakes, lizards, and insects. Because of this, temperature has a large influence on their physiological traits, or traits that effect their fitness, such as how quickly they develop or how long they live. This relationship between temperature and traits is often studied, as it can aid in developing effective control and mitigation strategies under changing climates. For example, if mosquitoes lay more eggs and develop quicker at certain temperatures, rising global temperatures could shift where these conditions occur, potentially expanding mosquito populations and the diseases they transmit. We can quantify this relationship using statistical methods and models to help predict these changes. However, there is a lot of ambiguity in how to do this and a deficiency of quality data to do so. The purpose of this work is to identify statistical methods that best capture the trait–temperature relationship in arthropods (a subcategory of ectotherms) while addressing factors that are often overlooked. We show how different assumptions about what the data look like affects our ability to make conclusions, use data from well-studied species to try to improve the accuracy for less-studied species, and present a method that will effectively include a second environmental variable (such as relative humidity) in addition to temperature into our models
Transcending the Fight: Culture, Power, and Adaptation to the Information Environment
No full textDoctor of PhilosophyIn 2007, the world saw the iphone, a revolutionary new product that connected individual people to the world wide web, giving them instantaneous access to each other and accelerating the digitization of the world. Malicious actors are now using this digitized world to project information power in order to get what they want, regardless if they are using unethical or illegal means such as cyber hacking or creating disinformation campaigns to skew voters' opinions.
This poses a new challenge for national security professionals as they struggle with how to perceive the use of military force against adversaries that operate in this space. While they use cyber and disinformation against the US military, they are increasingly using those tools against the American people. This new form of power projection is growing in its use and effectiveness, disrupting basic infrastructure services, economic freedom, and distorting perceptions of what is real and what is not. Some countries are adapting to this new form of power projection better than others. Why is that?
This research seeks to answer that question. It looks at government power and strategic culture as possible explanations for adapting to this new informatized world. The dissertation looks at the United States, China, Russia, and the United Kingdom and studies if they are successfully adapting to this new information environment and, if so, why? It concludes that culture is the largest factor and, for the United States, the culture of liberty (specifically the concern over individual rights) is the key reason why it has been the least adaptive nation
Investigating Seamless Transitions Between Immersive Computational Notebooks and Embodied Data Interactions
No full textA growing interest in Immersive Analytics (IA) has led to the extension of computational notebooks (e.g., Jupyter Notebook) into an immersive environment to enhance analytical workflows. However, existing solutions rely on the WIMP (windows, icons, menus, pointer) metaphor, which remains impractical for complex data exploration. Although embodied interaction offers a more intuitive alternative, immersive computational notebooks and embodied data exploration systems are implemented as standalone tools. This separation requires analysts to invest considerable effort to transition from one environment to an entirely different one during analytical workflows. To address this, we introduce ICoN, a prototype that facilitates a seamless transition between computational notebooks and embodied data explorations within a unified, fully immersive environment. Our findings reveal that unification improves transition efficiency and intuitiveness during analytical workflows, highlighting its potential for seamless data analysis.Published versio
Computational Analysis and Network-based Modeling of Cross-Species Transmissions
No full textZoonotic spillover of pathogens is the dominant cause of emerging infectious diseases.Cross-species transmission (CST) risks are accelerated by climate change, which alters animal habitats and aggregates new combinations of host species at high population density and elevations. Modeling CST dynamics is essential in ecology and computational epidemiology to enhance preparedness and resilience against future outbreaks. However, accurate prediction remains challenging due to biased pathogen sampling in existing CST databases and complex interactions among viral host range.
This dissertation addresses three main challenges: (1) optimizing CST testing set selection through graph entropy frameworks; (2) modeling infectious pathways and biodiversity shifts with climate change scenarios; and (3) developing an accessible knowledge-based question and answering (QA) framework using Retrieval Augmented Generation (RAG) technology.
Current viral databases consist mostly of pathogens in humans and domesticated animals, while the remaining vertebrate genera account for a mere 9%. Testing resources are limited for assessing indeterminate 800,000 to 1.5 million mammalian viruses with zoonotic potential.
Furthermore, climate change will expose host species to novel ecological interactions and complicate efforts to identify infectious pathways. I leveraged information theory and graph entropy, where high entropy implies a more informative, diverse, and unpredictable network structure, to guide testing set selection, aiming to maximize the entropy and improve diversity of CST database.
A graph representation constructed based on animal habitats, climate classification, and future climate scenarios, identifies biodiversity patterns in climate classifications and vulnerable hosts and viruses. Lastly, this dissertation introduces a knowledge graph-based CST information system for question answering (QA) using RAG, comparing multiple external database architectures, including Knowledge graphs, node embeddings, and vector databases. The evaluation framework integrates reasoning, summarization, and hallucination detection using curated unanswerable queries.
Through computational modeling and graph-based analysis of CSTs, it identifies potential missing links and delivers an accessible and accurate CST information framework, facilitating early detection of CST risks and improving preparedness for future emerging infectious diseases.Doctor of PhilosophyThe majority of emerging infectious diseases (EIDs) that affect humans originate from non-human animals. The process known as zoonotic spillover, or cross-species transmission (CST), happens when pathogens are transmitted from one species to another. Climate change increases the risk of CST by forcing species into overlapping habitats and increasing contacts in certain regions, such as high elevation or highly dense areas. These environmental shifts the probability of viral spillover events and complicate global efforts to predict and mitigate future outbreaks.
This dissertation focuses on developing computational methods to model and analyze CST dynamics, addressing three major challenges in data imbalance, bias, and incompleteness of current CST data sets. It integrates information theory and graph entropy, knowledge graphs (KGs), and retrieval augmented generation (RAG) to improve graph diversity, optimize testing strategies, and provide accessible and accurate information.
There are three major research challenges in this work: 1) optimizing CST testing set selection using entropy based graph framework to maximize information gain, 2) construct heterogeneous graphs to model biodiversities per climate regions, infectious path changes under climate change scenarios, climate change scenarios, hosts, and viruses, and 3) developing question answering (QA) with RAG technique in the domain of CST.
These computational approaches provide a scalable framework to understand the impact and potential risks in zoonotic emergence. Through modeling and computational analysis in CSTs, this research aims to enable early detection of zoonotic risks, raise preparedness for future outbreaks, and generate more trustworthy information for infectious diseases
Navigation, the Journal of the Institute of Navigation
No full textA tightly coupled GPS/IMU estimation algorithm is developed assuming that all received measurements must first be authenticated by CHIps MEssage Robust Authentication (CHIMERA). CHIMERA is designed to authenticate incoming GPS signals through two methods referred to as the fast and slow channels. This paper analyzes the accuracy of estimation algorithms for both of these channels when using an Inertial Measurement Unit (IMU) to compensate for authentication delay, and it considers the effects of different quality IMUs. This paper also introduces a concept of authentication staggering as a possible approach to improve location and attitude accuracy. The estimation algorithm is modified to account for authentication staggering and different possible estimation architectures are developed for this purpose. The results indicate that the fast channel produces typical GPS navigation accuracy for different quality IMUs while the slow channel has moderately degraded navigation accuracy even with a navigation-grade IMU and highly degraded accuracy with tactical- and MEMS-grade IMUs. Staggering the authentication times of the GPS satellites can be used to improve navigation accuracy for the slow channel.Submitted versio
Diagnostics
No full textBackground/Objectives: Older adults are disproportionately affected by traumatic brain injuries (TBIs), representing a significant portion of TBI-related hospitalizations and deaths. The objective of this study was to evaluate the feasibility and effectiveness of BrainCheck (Braincheck, Inc., Austin, TX, USA), a digital cognitive assessment tool, in detecting acute TBI-related cognitive deficits in the context of dementia-related cognitive impairment in older adult emergency department (ED) patients.
Methods: From March 2020 to November 2023, participants aged 65+ with mild TBI, as defined by the American College of Rehabilitation Medicine (ACRM) diagnostic criteria, and individuals with isolated orthopedic injuries were recruited from 14 U.S. type 1 and 2 trauma centers. After informed consent, each subject was assessed by well-validated neurocognitive tests to characterize pre- and postinjury cognitive status. The Clinical Dementia Rating (CDR) and Functional Activities Questionnaire (FAQ) were used to assess cognitive impairment, with the informant sections used to classify preinjury status. The Rivermead Post-Concussion Symptoms Questionnaire (RPQ16) was used to assess injury-related symptoms, and the tablet-based BrainCheck Battery was tested as a diagnostic platform for injury-related deficits across several functional domains. Spearman's correlation was used to assess BrainCheck's internal validity and its relationship with self-reported cognitive symptoms. Technology familiarity was self-reported on a 1 (lowest) to 5 (highest) Likert scale. ROC curves evaluated the tool’s accuracy in identifying cognitive impairment associated with TBI in the context of pre-existing cognitive impairment.
Results: For the 101 mTBI and 52 orthopedic trauma control patients, BrainCheck demonstrated strong internal validity, with significant correlations among its component tests, indicating its effectiveness in assessing cognitive impairment. However, low correlations with RPQ16 self-reported symptoms suggest that BrainCheck and the self-reported questionnaire assess different aspects of cognitive functions.
Conclusions: While BrainCheck effectively identified cognitive impairment, the composite battery and scoring did not differentiate TBI and dementia. Technology familiarity did not affect test outcomes. BrainCheck is a useful tool for evaluating cognitive function in adults aged ≥ 65 years with and without TBI in ED settings.Published versio
Promoting Student Learning by Automatically Managing Resubmission Tokens
No full textIn many university computer science courses, strict deadlines can create stress and penalize students who face unexpected challenges. A popular alternative is to give students a few "late passes", "resubmission passes", or "tokens" to use on assignments throughout the semester. While this approach is more flexible and equitable, it creates a significant amount of manual work for instructors, who have to track every request and update deadlines across multiple websites. This thesis tackles that problem in two ways. First, it introduces the EGP Broker, an LMS-integrated LTI 1.3 tool that automates the entire process, removing the need for instructors to manually manage tokens or for students to perform extra steps to use them. Second, this thesis analyzes what happened when a large introductory computer science course switched from a traditional late penalty (a 10% deduction per day) to a manually managed flexible token resubmission based policy. The results indicate that the flexible policy was associated with higher pass rates, improved final course and exam grades, and reduced dropout rates. Most students thrived under this system, using the flexibility to manage their time effectively without compromising the quality of their work. While a small number of students appeared to struggle with using the policy productively, the overall impact was positive. Overall, flexible token policies show great promise in promoting student success, while also highlighting the importance of providing additional support for students who may need more structure.Master of ScienceStrict assignment deadlines in university computer science courses can create stress and can unfairly penalize students who face unexpected challenges. One increasingly popular alternative is to give students a small number of "tokens" that allow them to submit assignments late or revise their work. While these policies offer greater flexibility and fairness, they often require instructors to manually track requests and update deadlines across multiple online course platforms, creating a significant administrative burden. This thesis addresses both the practical and educational aspects of flexible deadline policies. First, it introduces a software tool that connects directly to a course's online systems and automatically manages the use of tokens, removing the need for instructors to handle requests by hand and allowing students to use tokens without extra steps. Second, the thesis examines the effects of replacing a traditional late-penalty policy with a manually managed, token-based resubmission policy in a large introductory computer science course. The results show that the flexible policy was associated with higher pass rates, better final course and exam grades, and lower dropout rates. Most students used the added flexibility to manage their time more effectively without lowering the quality of their work. Although a small number of students struggled with the increased independence, the overall impact was positive. These findings suggest that flexible token-based policies can support student success while highlighting the importance of additional guidance for students who benefit from more structure
Biomedical Optics Express
No full textThe transport of intensity equation (TIE) is a powerful phase imaging technique. However, its formulation fails to account for the off-axial transfer function due to the paraxial approximation. To address the resulting image degradation, we analyze TIE phase retrieval using the contrast transfer function (CTF) framework. The attenuation of high-frequency components, leading to the loss of fine structural details, is assessed and restored through Wiener deconvolution. Simulations and experiment results demonstrate significant enhancements in sharpness, contrast, and structural clarity. We showcase enhanced phase imaging of cheek cells, revealing finer subcellular details and achieving diffraction-limited performance, contributing to advances in super-resolution phase imaging.Published versio
"You Have to Focus on School…You Can't Focus on Getting Paid": A Multi-Study Exploration of Perspectives and Experiences of Engineering Students Who Work
No full textWorking while attending college is a prevalent practice among undergraduate students in the United States. However, adapting engineering programs to this reality and promoting the retention of undergraduate students remains a challenge. This research on working engineering students is significant as it sheds light on the need for such adaptation and its potential impact on engineering education.
In this multi-study dissertation, I highlight the realities of working engineering students across multiple contexts, including community colleges and four-year universities, from various perspectives, such as those of students and advisors, in three manuscripts. The first manuscript is a phenomenologically informed qualitative study that focuses on the experiences of community college students and their understanding of computing identity as it relates to their jobs and coursework in introductory artificial intelligence (AI) courses. Upon analyzing their interviews, I found that working while enrolled in introductory computing courses provides participants with opportunities to demonstrate their computing knowledge in the workplace, gain recognition as a computing professional, and develop their interest in computing.
The second manuscript is a qualitative study that focuses on the experiences of ten undergraduate engineering students who work while enrolled at an R1, 4-year university and how they navigate undergraduate engineering education while working. I studied the interactions between work and school in the experiences of engineering students who work, as well as the strategies they used to navigate their engineering education. I found that when participants encountered scenarios in which engineering education affected their work, they modified their work responsibilities during the academic year. When they faced situations where work affected their engineering education, they often avoided out-of-class engagement opportunities or modified their course schedules or routines (e.g., dropped classes, took semesters off).
In Manuscript 3, I examined the elements of engineering education at universities that most frequently negatively impact engineering students who work, and how these elements affect them. For this study, academic advisors from nine R1 universities responded to a questionnaire that inquired about aspects of engineering education, drawing on the literature on students who work. Using descriptive statistics and descriptive qualitative coding, I analyzed the questionnaire responses. I found that the elements that most frequently negatively affected engineering students who work included time-intensive assignments, course scheduling, and teamwork expectations. These elements impacted students by increasing their time to graduation, reducing their engagement in extracurricular activities, and negatively affecting their academic performance.
This research on working engineering students not only highlights the complexities in their situations but also offers practical implications for addressing these challenges. The solution is not necessarily to advise students to stop working, but to find ways to reconcile the tensions between their roles as students and workers. Implementing such strategies could benefit both students and institutions.Doctor of PhilosophyWorking while attending college is common among undergraduate students in the United States, yet adapting engineering programs to support working undergraduate students remains a challenge. This dissertation highlights the experiences of working students in various contexts and from multiple perspectives. In this multi-study dissertation, I examine the experiences of working engineering students across community colleges and four-year, research-intensive universities through three manuscripts. The first manuscript focuses on the experiences of community college students with computing identity in introductory computing courses. The findings revealed that participants' jobs enhance their computing knowledge, recognition, and interest in the field. In the second manuscript, I investigate the challenges faced by ten undergraduate engineering students at a four-year research-intensive university as they balance work and studies. The findings indicated that when their engineering education conflicted with job responsibilities, students often adjusted their work commitments or modified their academic engagement, such as dropping classes or avoiding extracurricular activities. In the third manuscript, I distributed a questionnaire to academic advisors at 9 four-year, research-intensive universities to identify elements of engineering education (i.e., course scheduling and sequencing) that negatively impact working students. The findings indicated that time-intensive assignments, course scheduling, and teamwork expectations more often hindered these students, resulting in decreased academic performance and longer graduation times. Combined, the findings from these studies highlight the challenges faced by working engineering students and provide practical solutions to help them balance their dual roles as students and workers. This work emphasizes that the objective should not be to advise students to stop working, but rather to implement strategies that enhance their success without compromising their employment. By addressing these challenges, institutions can better support working engineering students and foster an environment conducive to their academic and professional growth
Social Science
No full textThis study analyzes the perspectives of support providers to survivors of child sexual abuse (CSA) on the potential links between pornography and the sexual abuse of children. Drawing from fifty interviews, eight focus group discussions, and post-interview surveys with frontline child advocacy support professionals from various backgrounds and settings, each with at least five years of experience in the field, this paper presents a conceptual model that situates pornography and CSA within interconnected “zones of violence” across digital, institutional, and community environments. Participants identified overlapping risk factors that can heighten pornography exposure and CSA vulnerability, including strained guardian–child relationships, inadequate supervision and digital literacy, socioeconomic precarity, limited access to services, and restrictive or patriarchal sexual norms. They described mediating processes linking pornography to abuse—social modeling, normalization of coercive and violent sexual scripts, grooming, power/threat dynamics (including sextortion and blackmail), and the production and circulation of child sexual abuse material (CSAM). Respondents perceived pornography as pervasive in young people’s lives, reported that it contributes to perceived shifts in CSA patterns, and emphasized the absence of best practices. They advocated comprehensive, digitally literate sex education; routine, developmentally appropriate screening; trauma-informed responses that avoid labeling and criminalizing children; and coordinated, multidisciplinary reforms.Published versio