ETD -- Doctoral Dissertations

Permanent URI for this collection

https://jscholarship.library.jhu.edu/handle/1774.2/838

For information about submitting electronic theses and dissertations, please see the ETD information page.

Browse

Now showing 1 - 20 of 5503

POLIOVIRUS VACCINATION IN THE CONTEXT OF THE  COVID-19 PANDEMIC: AN EXPLORATION OF WHETHER AN ESTABLISHED VACCINE COULD HAVE VALUE IN FIGHTING A NEW AND EVOLVING VIRUS, AND THE IMPACT OF BIOLOGICAL AND SOCIAL FACTORS ON THIS STRATEGY
(Johns Hopkins University, 2023-07-17) Comunale, Brittany A.; Celentano, David; Engineer, Lilly D; Klein, Sabra L; Hsu, Yea-Jen; Larson, Robin J; Marsteller, Jill; Engineer, Cyrus
Theory: Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the virus that causes coronavirus disease 2019 (COVID-19), emerged in 2019, rapidly spread across the globe, and continues to be transmitted today. Vaccination is the primary measure available for preventing infection and reducing severity of disease, however because currently available COVID-19 vaccines all target the virus' spike protein, which can evolve from one variant strain to the next, their efficacy can quickly wane. In order to stay ahead of future variants, researchers have sought to identify prophylactic agents (i.e., drugs and vaccines) that target proteins that are more highly conserved, less mutable. In addition, because proteins are often shared across similar RNA viruses, there is interest in whether drugs and vaccines developed for other RNA viruses could be effective against SARS-CoV-2. To help address these questions, using the well-established poliovirus vaccine as an example, this dissertation examines whether existing agents that target the highly conserved RNA-dependent RNA polymerase (RdRp) protein, could be repurposed to prevent and treat SARS-CoV-2, and how biological and social factors might impact such a strategy. Methods: The first manuscript presents a scoping review of the functional implications of RdRp in the context of the COVID-19 pandemic. Studies examining if and how existing drugs and compounds inhibit SARS-CoV-2 RdRp activity are summarized. The second manuscript examines vaccination histories of adults recently inoculated with the inactivated poliovirus vaccine (IPV) and the potential impact on SARS-CoV-2 infection and COVID-19 symptoms. The third manuscript further explores this study population, focusing on risk factors related to biological and social determinants of health. Results: The scoping review indicates that existing drugs and compounds that target the RdRp protein can not only disrupt SARS-CoV-2 replication of the original “Alpha” strain, but also subsequent variants. These findings validate the idea that agents targeting highly conserved proteins can be effective regardless of which variant is circulating, and suggest that drugs initially developed for a different RNA virus could have value when repurposed against SARS-CoV-2. Analyses of survey data from adults recently vaccinated with the inactivated poliovirus vaccine (IPV) reveal underlying medical conditions, employment status, vitamin D supplementation, and education level may impact subsequent COVID-19 outcomes. Additionally, regression models adjusted for vaccination histories as well as biological and social determinants of health show that oral poliovirus vaccination (OPV) is associated with a significantly lower incidence of SARS-CoV-2 infection and a shorter duration of COVID-19 symptoms. Conclusions: Prophylactic and therapeutic agents that target RdRp may complement current COVID-19 prevention and treatment tools. Clinical trial survey data support studies that suggest IPV can boost mucosal immunity in OPV primed individuals. Future, larger-scale studies are recommended.
DEVELOPING AND APPLYING A KNOWLEDGE TRANSLATION FRAMEWORK TO EVALUATE REMOTE CAPACITY BUILDING WORKSHOPS IN NIGERIA AND INDIA THROUGH THE STRIPE INITIATIVE
(Johns Hopkins University, 2023-07-17) Bala, Jay S; Frattaroli, Shannon; Davey-Rothwell, Melissa; Alonge, Olakunle; Closer, Svea; Bush, Seth
Background Workshops are a common Knowledge Translation (KT) tool in Public Health to build capacity in a relatively short period of time, especially for working professionals. This research aims to better understand the impact of remote methods on workshops and conduct an evaulation on impact of two remote workshops conducted as part of the STRIPE (Synthesis and Translation of Research and Innovations from Polio Eradication) initiative. These papers will do the following. First conduct a literature review to outline the strengths, weaknesses opporuntity and threats associated with using remote workshops and identify the best practices to mitigate those weaknesses and threats. Second, develop a new conceptual framework leveraging KT, learning evaluation and implementation science theories to generate a consolidated framework to evaluate multi-site workshops. Third, implement the tools generated from that framework to evaluate changes in knowledge and self-efficacy from the two workshops conducted as part of the STRIPE initiative. Methods The first manuscript will be conducted using a PRISMA literature review methodology followed by content analysis to develop themes for the analysis. The second manuscript will follow Jabareen’s methodology for generating conceptual frameworks and the third manuscript will implement a traditional pre-post survey to collect information about the workshop and the participants to evaluate the impact and efficacy of the workshop. Statistical comparisons will be done via the McNemar test. Results The first manuscript outlines that remote workshops are feasible and may have more advantages compared to in-person but there are unique pitfalls to a remote model including ensuring technology access, usability, and engagement. The main best practice identified was taking the time to appropriately plan, adapt and implement a remote workshop. The conceptual framework developed for evaulation was a blending of 24 concepts, classified over 2 levels of context and 5 steps of knowledge translation from engagement to behavior change. The two STRIPE workshops were evaluated using this framework, demonstrated modest gains in knowledge and self-efficacy; however, identifying co-variates was difficult due to sample size and lack of qualitative. The use of the novel framework provided a unique lens in evaluating the STRIPE’s multi-site workshops showing that adaptions and different implementation styles had an impact on the evaluation outcomes. Conclusions Overall, this body of research provides a practical guide for educators when planning remote workshops, a set of tools to evaluate those workshops and an example of how those workshops could be evaluated. Future research is required to validate the conceptual framework.
IMPACT OF INCLUDING ELECTRONIC NICOTINE DELIVERY SYSTEMS (ENDS) IN A STATE INSPECTION PROTOCOL ON STATE RETAIL VIOLATION RATE IN THE SYNAR YOUTH TOBACCO PREVENTION PROGRAM
(Johns Hopkins University, 2023-06-20) Hagen, Clark; Celentano, David; Engineer, Lilly; Morlock, Laura; Xie, Anping; Jacobs, Megan
Introduction and Background: The Synar regulation (45 C.F.R. § 96.130) is a Federal regulation and program administered by the States that requires enforcement to prevent the sale of tobacco products to those under the legal age of sale. Penalties are given to States that exceed a federally mandated retail violation rate (RVR). Tobacco use is changing with youth use of cigarette smoking decreasing and use of Electronic Nicotine Delivery Systems (ENDS) is increasing. There is a high economic and personal cost of ENDS use, although the full range of long-term negative outcomes of ENDS use are yet to be quantified. States are beginning to include ENDS in their Synar inspections and may worry that inclusion of ENDS may reduce their RVR compliance. Methods: RVRs were collected and listed from each of the 50 States and District of Columbia through manually reviewing each State’s 2020 RVR. Retail Violation Rate was used to compare with inspection of ENDS, presence of a high prevalence of ENDS use in the environment, and presence of regulations of ENDS. Data was collected for State RVR, Retail Availability of ENDS, ENDS Use in State, Use of Under 21 Year Old Inspectors, ENDS Included in Inspections, Inspectors Carry ID, Any Change in Law Regarding Inspections, and Other Changes in State Law. RVRs were dichotomized into categories of high RVR and low RVR and analysis was conducted. Results: Inclusion of ENDS in Synar inspections was not found to have significant impact on RVR. High retail availability and youth inspectors carrying identification had a low association with a State having a high RVR. High ENDS use in a State, changes in state law around inspections, other state law changes, and use of inspectors over the age of 18 were not found to be related to high RVR. Discussion: A State should not be concerned that including ENDS in its Synar inspection protocol will have impact on State compliance with the requirements of the Federal Synar Regulation. These results can allow States to best target prevention resources based on the findings of which variables influence RVR. Best practices to account for these variables can be identified and targeted policies and regulations can be implemented.
THE ROLE OF PUBLIC HEALTH POLICY IN SUPPORTING STATE AND LOCAL CONTACT TRACING PROGRAMS IN THE UNITED STATES
(Johns Hopkins University, 2023-07-03) Woodward, Alexandra; Sharfstein, Joshua; Rivers, Caitlin
Case investigation and contact tracing (CI/CT) are fundamental public health strategies that have been used to break chains of infectious disease transmission in a variety of historic outbreaks. It is critical that public health agencies have CI/CT program components in place prior to the detection of an outbreak to ensure that these programs are ready to be leveraged when an outbreak is first detected. The goal of this research is to identify the capacities and capabilities required to implement and sustain effective CI/CT programs in United States (U.S.) state, tribal, local, and territorial public health agencies (herein, public health agencies) and to identify key federal policies and resources needed to support these programs so they can be readily leveraged in future outbreaks. This research was conducted in three parts. First, a narrative literature review was conducted to identify the capacities, capabilities, outcomes, and impacts of CI/CT programs. Second, semi-structured qualitative interviews were held with ten state and local public health agencies as well as four experts in CI/CT to discuss the capacities and capabilities identified during the literature review and lessons learned around scaling-up and sustaining CI/CT programs during the COVID-19 pandemic and the 2022 mpox outbreak. Finally, a narrative literature review and interviews with six public health policy experts explored federal policies that impacted the scale-up and maintenance of several CI/CT capacities during the COVID-19 pandemic and the 2022 mpox outbreak. Findings from the narrative literature review and interviews resulted in the first comprehensive documentation of CI/CT capacities, capabilities, outcomes, and impacts of CI/CT programs and a conceptual framework that illustrates the relationships between these components. The interviews explored the CI/CT capacities that should be maintained as baseline functions in public health agencies on a continuous basis to ensure preparedness to conduct CI/CT in future outbreaks, and identified lessons learned around scaling-up CI/CT during the COVID-19 pandemic and the 2022 mpox outbreak. The second narrative literature review and interviews with public health policy experts resulted in federal policy recommendations to ensure public health agency preparedness to conduct CI/CT during future outbreaks and pandemics.
PREDICTING SUICIDE: UTILITY OF SOCIAL DETERMINANTS OF HEALTH AND ACCESS DATA
(Johns Hopkins University, 2023-06-12) Castner, Matthew David; Bittle, Mark J.; Wilcox, Holly C.; Kharrazi, Hadi H.; Nestadt, Paul S.; Clarke, Diana E.
By the numbers, 45,979 people died by suicide in 2020, making suicide the twelfth leading cause of death overall in the United States and the second-leading cause of death for ages 10-34.1 Suicide remains a particularly difficult condition upon which to intervene. In this respect, it perhaps resembles another high-mortality disease, cancer. A commonly held belief about cancer is that it is a single disease, while the reality is that it is a hundred – or hundreds – of different diseases each with its distinct etiology and clinical course.4 While suicides bear a superficial resemblance in their result, they are the result of a hundred, or hundreds, different pathways. While several risk factors of death by suicide, such as a diagnosis of depression or firearm ownership, stand above the rest, it is largely a murky picture. Suicide Prediction Models (SPM) are one way to understand and trace those pathways. These statistical models use data about the facts of a person’s life to estimate their risk of suicidal behavior. SPMs are a particularly useful tool as they can consider a broad range of potential risk factors and boil that down to a single measure of risk. A common use case for these would be to identify patients at risk for death by suicide and prompt some kind of intervention. Models like this could also be used to guide interventions to reduce the population-level risk of suicide. Through this thesis, I attempt to illustrate the role of Social Determinants of Health (SDoH) and Access to Care play in improving the performance of SPMs. The first manuscript within the thesis explores the extant literature on this subject. The second manuscript describes the development of a SPM to test the value of SDoH data. The third manuscript details an ecological model to determine the impact of SDoH and Access data on predictive ability by SPMs.
INDIVIDUAL AND COMMUNITY-LEVEL FACTORS ASSOCIATED WITH TELEHEALTH UTILIZATION AMONG VULNERABLE MEDICARE BENEFICIARIES DURING THE COVID-19 PANDEMIC
(Johns Hopkins University, 2023-06-07) Baawuah, Prince; Weiner, Jonathan; Castillo Salgado, Carlos; Kharrazi, Hadi; Commodore-Mensah, Yvonne; Gaskin, Darrell
Background: The COVID-19 Pandemic increased the use of telehealth as an avenue for vulnerable populations to receive services while social distancing. Together with telehealth policy changes, the use of telehealth soared across the United States (U.S.) health system. However, access to these digitally supported services may not have been equitable across the U.S. Objective: Using claims data, the study documents the use of Medicare Part B telehealth-eligible services within the fee-for-service Medicare program during the first year of the Pandemic. It also identifies key individual and community level characteristics associated with how vulnerable and special needs persons utilized telehealth. Methods: This is a cohort study of continuously enrolled Medicare FFS beneficiaries in 2019 and 2020. The main outcome is the utilization of telehealth-eligible services. This study used descriptive analyses and a multivariable logistic regression. Results There were 6,233,884 total beneficiaries who used telehealth eligible services with 57% females. Mean [sd] age is 71.1 [11.27] years. Between 2019 and 2020, total visits decreased by 9%. But this reduction would be lower still, if it wasn’t that telehealth contacts served as alternatives for a significant number of in-person visits in 2020. Telehealth ambulatory utilization accounted for about 1% of contacts in 2019 but 10% in 2020. Individual level factors associated with higher telehealth utilization in 2020 include younger age, female, American Indian/Alaska Native status, not having both Medicare and Medicaid coverage, having a higher number of chronic conditions, and not having the End-Stage Renal Disease. Community level socioeconomic determinants associated with higher telehealth utilization include residing in New England, urban areas, areas with lower levels of internet connectedness, higher community vulnerability. These analyses also stratified across the disabled (under 65) and the aged (over 65) subgroups. The impact of some individual and community-level factors varied across the two subgroups. Conclusion: Individual and community level factors are associated with different levels of utilization of telehealth services. Disability status effect modifies the relationship between the various individual and community level characteristics and telehealth utilization. Therefore, all stakeholders should understand these differences to affect equitable telehealth policies and strategies across the diverse Medicare population.
AN ECONOMIC EVALUATION, ALONGSIDE A RANDOMIZED CONTROLLED TRIAL, TO ASSESS THE EFFECTIVENESS OF A CASE-FINDING STRATEGY FOR TUBERCULOSIS IN SOWETO, SOUTH AFRICA
(Johns Hopkins University, 2023-06-01) Bhardwaj, Vinayak; Dowdy, David; Patenaude, Bryan; Rao, Krishna; Salazar-Austin, Nicole; Martinson, Neil; Shet, Anita; Golub, Jonathan
Problem statement: Novel methods of improving the identification of patients with undiagnosed TB are required. A possible approach, assessed in the ACTTIS trial in South Africa, is to identify and link to treatment household contacts of paediatric patients with confirmed TB. However, the cost-effectiveness of this approach, the costs of scaling it up, and its impact on TB transmission overall, are not known. Methods: We used a cost-effectiveness model to compare the costs and effectiveness values of the intervention (ACTTIS arm) and the control (SOC+), using measured costs in ZAR and DALY values as a measure of effectiveness. The incremental cost-effectiveness ratio (ICER) values and various sensitivity analyses were computed. We calculated the total cost of scaling the intervention (and SOC+ control) to all paediatric index patients in South Africa and their household contacts, as well as costs averted through reduced hospitalization. We used a Susceptible Exposed Infected Recovered (SEIR) model to simulate the impact of introducing the intervention in a hypothetical South African population. Results: The overall ICER of the ACTTIS intervention relative to the SOC+ arm was computed as -ZAR 4,954 per additional DALY averted (95% CI: -5,148; -4,760). This is highly cost-effective. One-way and probabilistic sensitivity analyses showed robustness to various input parameters. The total budget impact of scaling up ACTTIS or SOC+ to all undiagnosed paediatric patients and their immediate household contacts was estimated to be ZAR 1.219 billion for ACTTIS and ZAR 644 million for SOC+, accounting for hospitalization costs. Our simulation model showed that over a 10-year period, implementation of ACTTIS in a hypothetical cohort of 100,000 South African patients resulted in a reduction of total infections by 45% (from 6,807 to 3,611) and averted 929 deaths (95% CI: 889, 973), with an ICER of ZAR 5,760.62 per death averted (95% CI: 5,500.96;6,021.11). The net benefits of the intervention approximate ZAR 1.3 billion per 100,000 population (ZAR 1.2-1.37 billion) at a return of investment of ZAR1.96 per ZAR 1 invested in the ACTTIS program (1.86-2.08). Conclusions: The case-finding strategy deployed in the ACTTIS trial is highly cost-effective, potentially reducing the transmission of TB.
An Emerging Approach to Health Equity Practice: Exploring the Implementation of Organizational Health Equity Capacity Assessments
(Johns Hopkins University, 2023-05-12) Marcus, Rachel; Bowie, Janice; Pollack Porter, Keshia; Resnick, Beth; Frattaroli, Shannon; Rubenstein, Leonard
Objective: Public health organizations play a key role in achieving health equity, but there are significant gaps in the literature related to the assessment of organizations’ health equity capacity. The purpose of this dissertation research was to strengthen the evidence base related to organizational health equity capacity assessments (OCAs). This research benefits health departments by providing strengthened evidence related to OCA selection and implementation, helping departments to better assess their organizational capacity to design, implement, fund, manage, evaluate, and sustain health equity-oriented work. Methods: This dissertation contains three manuscripts. The first manuscript is a scoping review characterizing the OCAs in the gray and peer-reviewed literature, providing a baseline for researchers and practitioners to find and utilize the OCA that best meets their needs. The second manuscript explores the factors that facilitate or inhibit OCA implementation, and documents the initial organizational impacts of these assessments, through two case studies conducted with the Kitsap Public Health District (KPHD) and the Rhode Island Department of Health (RIDOH). The third manuscript is a white paper exploring the programmatic opportunities for OCA implementation and recommends further research. Results: The scoping review identified and characterized 17 OCAs that met the inclusion criteria at the time of research. All identified OCAs assess organizational health equity readiness and/or capacity, but differ regarding thematic focus, structure, and intended audience. Implementation evidence is limited. The case study expanded this evidence base, providing implementation evidence drawn from the two health department OCAs that will be useful to other departments interested in assessing their capacity. Considerations for future OCA implementation are highlighted in the results. The white paper highlighted additional research needs to strengthen OCA impact and identified potential programmatic uses of OCAs including to strengthen equitable public health emergency preparedness, develop equity-oriented public health capabilities through accreditation, and facilitate multi-sectoral, collaborative progress towards improved health equity action. Conclusion: This dissertation advances the evidence base related to organizational health equity capacity assessments and identifies opportunities for OCA utilization and further research. Organizational health equity capacity is a unique type of capacity and should be an ongoing focal area for all health departments.
HARD, UNSTABLE, SELF-SERVING STALEMATES (HUSSS) AND NEGOTIATION COLLAPSE: THE VENEZUELAN EXPERIENCE (2014-2021)
(Johns Hopkins University, 2023-09-27) Vidaurri Mudd, Marian Lourdes; Vuković, Siniša; Zartman, William; Kocher, Matthew A; Power, Timothy J
Conflicts and crises in a situation of deadlock that are highly resistant to peacemaking beg the question of why negotiation collapse persists, especially if the objective is to prevent these crises from becoming intractable and normalized. This research focuses on the aspect of deadlock in conflicts and its role in the likelihood of negotiation breakdown. Using the six cases of collapsed negotiations to solve the Venezuelan crisis from 2014 to 2021, a variant of stalemate where ripeness is absent is proposed. It is argued that not all deadlocks necessarily lead to the breakdown of peacemaking efforts, but under certain conditions —the presence of a Hard (polarization), Unstable (fragmentation), Self-Serving Stalemate (HUSSS)— is likely to lead to negotiation collapse. The cases of negotiations analyzed show that the combination of the three conditions of HUSSS produces a pattern of collapse, which can be overcome if there are intentional efforts at addressing the polarization between negotiating parties, fragmentation within domestic and international actors, and the self-serving nature of the status quo that has made the prospects of a negotiated agreement consistently unappealing.
The microanatomy of human skin in aging
(Johns Hopkins University, 2023-07-24) Han, Kyu Sang; Wirtz, Denis; Wu, Pei-Hsun; Joshu, Corrine; Sofou, Stavroula; Sunshine, Joel
Skin is the largest organ in the body, provides critical barrier functions, and yet visibly ages with time. Some global morphologic change like skin wrinkling and increased fragility are evident; however, how ageing affects the skin at the micro-anatomical and cellular scales is poorly understood. Here, we developed a deep learning-based workflow to deeply characterize the microanatomical tissue and cellular features that change in human skin with age. We extract 1,090 objective structural features and identify 124 which are significantly affected by age. We identify eight biomarkers of human aging in skin, six of which were previously unknown, including: dermal thinning, decreased size/number of hair follicles and sebaceous glands, and progressive horizontal alignment of the extracellular matrix and stromal cells. These biomarkers allow significantly improved prediction of skin age over comparison. These rigorously curated and segmented atlases of normal skin microarchitecture constitute important reference maps for future studies in age-associated diseases of the skin.
Some Topics in Neural Network-based System Identification
(Johns Hopkins University, 2023-07-20) Cui, Tianqi; Kevrekidis, Yannis; Tryggvason, Gretar; Betenbaugh, Michael; Bukowski, Brandon; Fazlyab, Mahyar
Neural network-based system identification is a modeling technique that uses a neural network to learn the relationship between the input states and output states of a system. The neural network is trained on a set of input-output pairs, and once trained, it can be used to predict the system response to new inputs. Neural network-based system identification is particularly useful for modeling complex systems with nonlinear dynamics and unknown or time-varying behavior. It is commonly used in control engineering, signal processing, and robotics to model and control complex systems.
DATA ANALYSIS AND MACHINE LEARNING FOR ENHANCING RESILIENCE TO FIRE, FROM IGNITION MAPPING TO STRUCTURAL AND SYSTEMS MODELING
(Johns Hopkins University, 2023-07-19) Tong, Qi; Gernay, Thomas; Igusa, Tak; Rajaram, Hari; Tam, Wai Cheong; Pita, Gonzalo
Fire hazards pose significant threats to our communities. Mitigation of fire risk requires an understanding of a range of issues and processes at various scales, such as the occurrence of ignitions in a community, the performance of a building structure under fire, and the efficiency of prevention and protection strategies. In this thesis, we investigate several of these fire safety issues through the lens of data-driven methods. Bayesian methods and machine learning techniques are adopted and tailored to address selected fire hazards and provide contributions toward solving these challenges for a fire resilient built environment. The thesis focuses first on fire ignitions. It investigates the problems of fire following earthquakes and wildfires. These issues are studied at the scale of a city or a region. For fire following earthquakes, a hierarchical Bayesian method is developed to allow modeling with scarce data, while for wildfire ignitions, an ensemble-based machine learning model is adopted. Then, the thesis zooms in to the building scale to assess data-based methods for evaluation of structural fire performance. Surrogate models are derived based on machine learning to capture the capacity of slender steel members in fire. Finally, the thesis investigates a framework to assess system resilience under fire hazards. The framework is applied for resilience assessment of facilities subjected to fires in the process industry. The thesis applies different data-based and modeling approaches to deal with fire hazards from different perspectives and at different scales, with the aim to enhance fire safety and build a more resilient environment against fires for our community.
UNSUPERVISED SEGMENTAL MODELING OF SPEECH FOR LOW RESOURCE APPLICATIONS
(Johns Hopkins University, 2023-07-19) Bhati, Saurabhchand; Dehak, Najim; Moro-Velazquez, Laureano; Patel, Vishal; Villalba, Jesus
Voice-enabled interfaces for human-machine interaction have made significant progress in recent years. Most of the success can be attributed to deep neural networks trained on thousands of hours of transcribed data. However, vast amounts of labeled data are not available for most spoken languages worldwide, e.g., regional languages. Here we explore alternate techniques that can learn directly from data without any or minimal manual transcriptions. The representation techniques employed to characterize the speech signal dictate the performance of unsupervised systems. Self-supervised methods such as Contrastive Predictive Coding (CPC) have emerged as a promising technique for representation learning from unlabeled speech data. Based on the observation that the acoustic information, e.g., phones, changes slower than the feature extraction rate in CPC, we propose regularization techniques that impose slowness constraints on the features. First, we propose two regularization techniques: Self-expressing constraint and Left-or-Right regularization. Our modifications outperform the baseline CPC in monolingual, cross-lingual, or multilingual settings on the ABX and linear phone classification benchmarks. However, CPC or our modifications mainly look at the audio signal's structure at the frame level. The speech structure exists beyond the frame level, i.e., at the phone level or even higher. We propose a segmental contrastive predictive coding (SCPC) framework to learn from the signal structure at both the frame and phone levels. SCPC is a hierarchical model with three stages trained in an end-to-end manner. In the first stage, the model predicts future feature frames and extracts frame-level representation from the raw waveform. In the second stage, a differentiable boundary detector finds variable-length segments. In the last stage, the model predicts future segments to learn segment representations. Experiments show that our model outperforms existing phone and word segmentation methods on TIMIT and Buckeye datasets. In the last part, we explore knowledge distillation from text encoders (e.g., Roberta) to speech encoders in an unsupervised manner in a multimodal setting. Text encoders operate at the sub-word level, while speech encoders operate at a much smaller scale, i.e., frames. Our segmental framework allows us to downsample frames and generate sub-words. SCPC enables us to leverage pretrained text encoders in an audio-visual and audio-only setting. We show significant performance improvements on the audio-image retrieval and semantic similarity task.
MEASUREMENT AND MITIGATION OF MICROBIAL CONTAMINATION ON VARIOUS PERSONAL PROTECTIVE EQUIPMENT
(Johns Hopkins University, 2023-07-20) Lee, Sang Hyuk; Mac Gabhann, Feilim; Donohue, Marc; Karunasena, Enusha; Sofou, Stavroula; Jeong, Sang Moo; Schulman, Rebecca; Doloff, Joshua
Recent global health crisis has called for methods to measure and mitigate microbial contamination on various personal protective equipment. Particularly, this document focu sed on: the implementation of modified AATCCAATCC-100 with qRT qRT-PCR assisted absolute and relative quantification for quantifiable tracing of both antimicrobial and microbial behavioral properties; the comparison of pipette tip repurposing efficiencies among lab detergent, ozone, and CAP; and the prospective application of vapor hydrogen peroxide, ozone, and CAP for mask repurposing. Log reductions from modified AATCC - 100 were compared to identify time dependent antimicrobial were compared to identify time dependent antimicrobial properties from silver ion containi containing wound dressing samples. A ntimicrobial properties of wound dressing samples diminished as incubation days are increased for both PCR and cell viability assay, while d ata from qRT qRT- PCR generally produced lower standard deviation than that of culture method s, hence shown to be more precise. Complementary parallel analysis of samples using both methods better characterized antimicrobial properties of the tested samples. A p arallel analysis using classical methods alongside the application of relative quantifi cation displayed changes in expression of virulence related genes. Although molecular assays targeting specific virulence activities are needed to verify the change in activities, relative quantification efficiently provided insight into changechanges specific t o virulence in model organisms. A c ontamination evaluation protocol were outlined t o evaluate the efficacies of the following repurposing methods: washing wit h a common laboratory detergent, exposure of ozone vapor, and CAP. Efficacy was determined by turn over ratio and log reduction in detectable genomic material of the contaminated products via re real -time quantitative PCR (qPCR). Ozone at 14400 PPM * minute is fully optimized while CAP shows promising potential post optimization. The application of ozone, hydrogen peroxide, and CAP is further explored for mask repurposing. Although further experimentation with BFE is needed, minimal change in physical properties of post post-repurposed masks showed promising potential as non non-destructive repurposing methods.
PROGRAMMABLE M13 HYPERPHAGE DISPLAY OF DIVERSE ANIMAL TOXIN LIBRARIES TO DISCOVER MEMBRANE PROTEIN TARGETED THERAPEUTICS
(Johns Hopkins University, 2023-07-19) Hsiao, Meng-Hsuan; Zhu, Heng; Larman, Harry Benjamin; Gray, Jeffrey; Ostermeier, Marc; Dong, Xinzhong
Animal toxins and cysteine-reinforced miniproteins present underexplored landscapes for drug discovery due to their unique structural attributes, potent bioactivity, and selective targeting. Recognizing the value of these compounds for therapeutic strategies, we developed a high-throughput strategy for large-scale screening of these miniproteins. To meet the urgent requirement for large-scale screening of cysteine-reinforced miniproteins, we devised a high-throughput strategy targeting the discovery of innovative drug candidates. Central to this strategy is the construction of two distinct libraries: the 'Animal Toxin' library, assembled from Uniprot database resources, and the 'Metavenome' library, developed to expand the former through sequence homologous proteins within an extensive Metagenomic database. Using programmable phage display integrated with high-throughput oligonucleotide library synthesis, we encoded these libraries with chosen polypeptides, thereby representing a substantial fraction of the cysteine-rich toxins universe for protein-protein interaction studies. We optimized our phage display using a programmable hyperphage technique for M13 phage display. Our hyperphage system enables the fusion of ligands with all five copies of the P3 protein expressed on the phage surface. This polyvalent display system not only enhances binding avidity but also amplifies the sensitivity to detect lower affinity interactions. Moreover, we have coupled single-round screening with next-generation sequencing (NGS) to evaluate the binding of all library members simultaneously. This innovative combination allows for rapid and efficient identification of potential ligands for target membrane proteins and can even detect interactions with very rare members of a library that might otherwise be outcompeted in a typical multiple-round panning process of traditional phage display. As an initial demonstration of the utility of our platform, we focused on two distinct receptors, epidermal growth factor receptor (EGFR) and Mas-related G-protein coupled receptor member X4 (MrgprX4), which are implicated in cellular growth processes and pain/itch signaling, respectively. We rediscovered known ligands, identified novel binders, and provided insights into potential binding modalities of these ligands. Our findings demonstrate the potential of our libraries in discovering bioactive ligands and the prospects of these binders as scaffolds for novel therapeutics. Altogether, our platform offers a promising approach for membrane protein-targeted drug discovery and exploring the structural diversity of cysteine-rich miniproteins.
Robust Speaker Recognition using Perceptual and Adversarial Speech Enhancement
(Johns Hopkins University, 2023-07-19) Kataria, Saurabh; Dehak, Najim; Villalba, Jesus; Moro-Velazquez, Laureano; Elhilali, Mounya
In Automatic Speaker Verification (ASV), we determine whether the speaker in the test utterance is identical to the previously enrolled speaker. Deep learning has significantly improved ASV performance. However, it is still susceptible to external disturbances and domain mismatches. A standard solution is data augmentation, i.e., adding noise and reverberations in training data. We focus on developing pre-processing solutions that can be integrated with existing pipelines and advance empirical performance on state-of-the-art systems. For this, we pursue deep learning-based speech enhancement and develop solutions equipped with denoising, domain adaptation, and bandwidth extension (BWE). Existing speech enhancement solutions often lead to degradation in ASV performance, partly due to loss of speaker information. We propose using perceptual/deep features that leverage pre-trained models to handle this. We also prove the effectiveness of our denoiser by showing that it complements the missing noise class in the x-vector (training) data augmentation through ablation studies. We also improve the training data for telephony speaker verification, where it is a common practice to downsample higher-bandwidth microphone speech to lower sampling frequency and apply telephone codecs. We propose to replace this by learning a mapping using a deep feature-based CycleGAN. Our novel technique improves training data and complements the prior techniques, including data augmentation. To handle bandwidth mismatch, we pursue BWE with time-domain architectures. We develop competent Generative Adversarial Networks (GAN): supervised (conditional GAN) and unsupervised (CycleGAN). Our findings indicate that unsupervised learning can give close performance to supervised performance. We also pursue joint learning BWE schemes with domain adaptation. Finally, with our proposed Self-FiLM scheme, we leverage self-supervised representations to guide BWE models better in unknown environments. In conclusion, we provide evidence that speech enhancement can be used as a pre-processor for improving ASV. By testing on real data acquired from Speaker Recognition Evaluation challenges, we demonstrate the effectiveness of speaker-identity preserving denoisers. We also study the effectiveness of domain adaptation and Self-Supervised Learning to improve bandwidth extension. Our work opens research into a joint investigation of enhancement-related problems and better generative models to assist the x-vector.
Automating the Analysis and Improvement of Dynamic Programming Algorithms with Applications to Natural Language Processing
(Johns Hopkins University, 2023-07-20) Vieira, Timothy; Eisner, Jason; Smith, Scott; Chiang, David
This thesis develops a system for automatically analyzing and improving dynamic programs, such as those that have driven progress in natural language processing and computer science, more generally, for decades. Finding a correct program with the optimal asymptotic runtime can be unintuitive, time-consuming, and error-prone. This thesis aims to automate this laborious process. To this end, we develop an approach based on (1) a high-level, domain-specific language called Dyna for concisely specifying dynamic programs (2) a general-purpose solver to efficiently execute these programs (3) a static analysis system that provides type analysis and worst-case time/space complexity analyses (4) a rich collection of meaning-preserving transformations to programs, which systematizes the repeated insights of numerous authors when speeding up algorithms in the literature (5) a search algorithm for identifying a good sequence of transformations that reduce the runtime complexity given an initial, correct program We show that, in practice, automated search—like the mental search performed by human programmers—can find substantial improvements to the initial program. Empirically, we show that many speed-ups described in the NLP literature could have been discovered automatically by our system. We provide a freely available prototype system at https://github.com/timvieira/dyna-pi
The Coevolution of Topography and Runoff Generation in Humid Landscapes
(Johns Hopkins University, 2023-07-14) Litwin, David G; Harman, Ciaran J; Rajaram, Harihar; Tucker, Gregory E; Barnhart, Katherine R; Lewis, Kevin; Viete, Daniel
Topography is an important control on runoff generation, as slope and relief affect hydraulic gradients, and curvature affects the convergence or divergence of flow paths. Over long timescales, however, runoff also shapes topography through surface erosion. This coevolution suggests that there may be a close relationship between landscape hydrology and topography that could provide insights into both hydrological and geomorphic processes. However, we do not have a strong theoretical framework for how topography and runoff generation should be linked, nor have there been many studies to determine how these links are expressed in the field. Here I address these areas, focusing on humid climates where runoff is primarily generated through groundwater return flow and precipitation on saturated areas. First, I present a new coupled model of runoff generation and landscape evolution that incorporates fluvial erosion driven by runoff from a shallow aquifer, hillslope diffusion, and uplift. Then, I nondimensionalize the model under the condition of steady and uniform groundwater recharge, and provide a mathematical framework for understanding the link between hillslope length, geomorphic process rates, and subsurface hydrological properties. Next, I explore the hydrological function of coevolved landscapes in more detail, focusing particularly on the emergence of variable source area hydrology. For this aim, I extend the model and nondimensionalization to include evapotranspiration and a simple representation of the vadose zone. I show, among other things, that coevolution with subsurface hydrology can explain why steeper landscapes are likely to have smaller variably saturated areas than landscapes with more gentle topography, and link this difference to subsurface properties and climate. Lastly, I test some of the model predictions in the field by exploring the hydrologic and geomorphic differences between two small watersheds in the Piedmont physiographic province near Baltimore that have contrasting subsurface architecture. I show that the site with a thin permeable subsurface has larger variable source areas and shorter hillslopes than the site with a thick permeable subsurface, as predicted by the model. A full parameterization of the model for the two sites suggests that subsurface properties are necessary to explain these differences.
DEVELOPMENT OF NEW METHODS FOR MATERIALS SCIENCE: TO ACCELERATE SIMULATIONS AND AUTOMATE ANALYSIS OF EXPERIMENTAL DATA
(Johns Hopkins University, 2023-07-18) Wang, Yunzhe; Paulette, Clancy; Mueller, Timothy K; Wang, Chao; Oses, Corey; Cheng, Lan
High-throughput simulations and characterization are essential for accelerating material development and reducing lab-to-market time. Considerable efforts have been devoted to the development of new algorithms and techniques to reduce the computational costs of materials simulations and streamline the quantitative analysis of experimental data, leveraging the latest advances in theories of material computation and machine learning paradigms. This thesis exemplifies such efforts in three projects. The first project develops a set of algorithms for the dynamic generation of optimized generalized Monkhorst-Pack k-point grids, which were shown to be capable of greatly reducing the computational costs of electronic structure calculations of crystalline materials. The new set of algorithms reduces the computational overhead in the search for efficient grids and eliminates the need for a pre-generated database by transforming the problem from an enumeration of three-dimensional superlattices to an enumeration of two-dimensional superlattices and a small set of symmetry-permitted shifts. A lightweight C++ library with a Python interface and a stand-alone Java application are developed for integration with existing DFT software. In the second project, a program is developed to accelerate the structure search of stable nanoclusters using machine-learned interatomic potentials and active learning. Interatomic potentials are updated on-the-fly during the exploration of the configuration space with DFT calculations on extrapolating structures and low-energy clusters in the genetic pool at each retraining stage, which guarantees accuracy of the final reported structures. Using this approach, new lowest-energy isomers of elemental aluminum clusters were found for 25 out of the 35 studied sizes. The third project involves developing a program to automate the characterization of X-ray Phase Contrast Imaging (XPCI) of particles emitted from metal composite combustion. The program automates the laborious manual characterization process and provides statistical analysis of particles. It consists of three stages: particle detection, trajectory reconstruction, and property analysis. Gaussian mixture model and convolutional neural networks are evaluated for particle detection, while the Kalman filter is used for trajectory reconstruction. A shape detection procedure classifies particles based on their shapes, leveraging classic computer vision algorithms. Finally, a graph data structure is employed to identify complex events like microexplosions and collisions.
POSE ESTIMATION AND SHAPE ABSTRACTION FROM A PROBABILISTIC AND GEOMETRIC PERSPECTIVE
(Johns Hopkins University, 2023-07-17) Liu, Weixiao; Chirikjian, Gregory; Cowan, Noah; Kazhdan, Michael
In recent years, we have witnessed great progress in how intelligent machines perceive the environment. To humans, perception is understanding the surroundings by interpreting sensing (e.g. vision, haptics, language) into semantic primitives, which are logistically understandable and mutually communicable. For machines, in their current form, perception can be defined as a process of formalizing a mathematical representation of the environment, from which higher-level tasks can be efficiently planned, modeled, and solved. Nowadays, most artificial intelligence (AI) systems rely on data-driven neural networks to establish versatile implicit representations. However, the learned representations are, to a large degree, task-specific, cumbersome, and weak in generalizability. Moreover, a learned representation is closed within itself, i.e., rarely interpretable and communicable to human beings or other AI systems. This dissertation focuses on achieving interpretable visual perception through geometric analysis and probabilistic modeling. In achieving the goal, this thesis targets two fundamental visual perception problems: pose estimation and shape abstraction. For the first problem, the author proposes a geometry-guided probabilistic model to realize an accurate and robust point cloud registration, which is the basis for understanding the pose relationship between visual inputs. Furthermore, the author proposes an unconstrained optimization method on a matrix Lie group to estimate the rigid transformation efficiently. Next, the methodology is extended to extract compact geometric primitive-based abstractions from visual inputs. This task goes beyond pose estimation and further includes the shape-level description of the visual input. Geometric primitives called superquadrics are studied and utilized as the basis for perceiving the environment. The primitive-based shape abstraction is concise and fully analytical in mathematics, which makes it a desirable standard messenger connecting perception and higher-level tasks such as segmentation, robot grasping, collision detection, motion planning, and physical simulation. Furthermore, the approach is learning-free, inherently generalizable, and interpretable, yet still possesses competitive expressiveness compared to learning-based representations.

Browse

Recent Submissions