UTSA Runner Research Press (Univ. of Texas at San Antonio)
Not a member yet
6846 research outputs found
Sort by
Examining the Impact of Natural Ventilation versus Heat Recovery Ventilation Systems on Indoor Air Quality: A Tiny House Case Study
Get PDFAdverse health effects can arise from indoor air pollutants, resulting in allergies, asthma, and other respiratory problems among occupants. Concurrently, the energy consumption of residential buildings, particularly concerning heating, ventilation, and air conditioning (HVAC) systems, significantly contributes to global energy usage. To address these intertwined challenges, heat recovery ventilation (HRV) has emerged as a viable solution to reduce heating and cooling demands while providing fresh ventilation rates. This study aims to investigate the indoor air quality (IAQ) of an experimental tiny house building equipped with an HRV unit by simulating real-life scenarios contributing to IAQ. The research evaluates the effectiveness of HRV compared to natural ventilation in managing particle matter (PM), total volatile organic compounds (TVOC), formaldehyde (CH<sub>2</sub>O), carbon monoxide (CO), and carbon dioxide (CO<sub>2</sub>) levels. This research significantly contributes to the understanding of the different ventilation strategies&rsquo; impact on IAQ in tiny houses and offers valuable insights for improving living conditions in a unique building typology that is underrepresented in the research literature.Architecture and PlanningElectrical and Computer Engineerin
Polar Region Sea Ice Classification With Multimodal Learning
Get PDFThe rapid changes in the polar regions due to climate change have significant global implications, necessitating accurate monitoring and classification of sea ice. This thesis presents a multimodal learning approach for classifying polar sea ice thickness using both satellite imagery and altimetry data. The objective is to distinguish between thick ice, thin ice, and open water, which are crucial for understanding ice dynamics and predicting future changes. Our methodology leverages deep learning models to extract meaningful features from satellite images and integrates them with complementary information from altimetry measurements, thereby enhancing classification accuracy. The proposed system employs a multimodal deep learning technique from a convolutional neural networks (CNNs) for polar image data and a RNN architecture for text data fusion, allowing the model to learn from multiple data modalities effectively. The experimental results demonstrate that the multimodal approach significantly outperforms single-modality models in terms of classification precision and accuracy. This work uses three main techniques to perform the fusion of the two modalities: LSTM model and U-Net model using Additive Fusion (AF), Multiplicative Fusion (MLF), and Gated Fusion (GF). Among the three techniques, GF outperforms the other two in terms of accuracy, precision, F1-score, and Recall.Computer Scienc
CPS Energy 2021 Annual Permit: Final Report for Ten CPS Energy Projects, Bexar County, Texas
Get PDFFrom July 27, 2021 through January 5, 2023, the University of Texas at San Antonio (UTSA) Center for Archaeological Research (CAR), in response to a request from Adams Environmental, Inc. (AEI), conducted cultural resources investigations on 10 project areas for CPS Energy (CPS). Because CPS is owned by the City of San Antonio (COSA) and is defined as a political subdivision of the State of Texas, the projects require review by the Texas Historical Commission (THC) under the Antiquities Code of Texas. CAR obtained an annual permit, Texas Antiquities Permit (TAP) Number 30154. Cynthia Munoz served as the Principal Investigator and Sarah Wigley, Peggy Wall, Jonathan Paige, and Leonard Kemp served as the Project Archaeologists.
The 10 archaeological investigations were conducted in advance of the installation of a gas main, utility poles, and substation infrastructure. They consisted of five intensive survey projects with shovel testing, one intensive survey project with shovel testing and backhoe trenching, and four cultural monitoring projects. Four new sites, 41BX2480, 41BX2481, 41BX2482, and 41BX2528 were recorded on two project areas, Whisper Falls and Howard Road Parcel 345. CAR recommends site 41BX2528 on the Howard Road Parcel and the portions of sites 41BX2481 and 41BX2482 within the Whisper Falls linear project alignment as ineligible for listing in the National Register of Historic Places (NRHP) or for designation as a State Antiquities Landmark (SAL). No further work is recommended for the three sites. CAR recommends the portion of site 41BX2480 within the Whisper Falls linear project alignment as having undetermined eligibility for listing in the NRHP or designation as a SAL due to moderately dense, deeply buried deposits and preservation of organic material suitable for radiocarbon dating. Because additional testing is necessary to make an eligibility determination, CAR recommends avoidance of the site. To comply with CAR’s recommendations for 41BX2480, CPS planned boring methodology for installation to successfully avoid impacting deposits associated with the site.
No materials were collected as part of these investigations. Associated records generated during this project are curated at CAR in accordance with the THC guidelines under CAR Accession 2742.Adams Environmental, Inc
Judgment, shame, and coercion: the criminal legal system and reproductive autonomy
No full textBackground
A growing body of research has called attention to limitations to reproductive autonomy in both women who are socially disadvantaged and in those who have had contact with the criminal legal (CL) system. However, it is unclear whether CL system contact influences contraceptive use patterns and how these processes unfold. We utilize a mixed-methods approach to investigate whether history of arrest is associated with receipt of contraceptive counseling, use of long-term contraception, sterilization, and subsequent desire for reversal of sterilization. We further consider how agents in and around the CL system may influence women’s reproductive decisions and outcomes (856 survey respondents; 10 interviewees).
Results
We observe that women who have been arrested more commonly report receipt of contraceptive counseling and sterilization. They are also significantly more likely to want their sterilization reversed. Our in-depth interviews suggest that women with CL contact experience considerable shame, and in some cases, coercion to limit fertility from various agents in and outside the criminal legal system including medical providers, Parole/Probation Officers (POs), guards, and family members.
Conclusions
Our findings suggest the need for ongoing attention to how exposure to this system may promote uneven use of certain forms of contraception and dissatisfaction, i.e., desire for reversal of sterilization, among these women. Findings further suggest that de-emphasizing the CL system as a means through which to address reproductive needs should be considered.Sociology and Demograph
Electrospun composite-coated endotracheal tubes with controlled siRNA and drug delivery to lubricate and minimize upper airway injury
No full textEndotracheal Tubes (ETTs) maintain and secure a patent airway; however, prolonged intubation often results in unintended injury to the mucosal epithelium and inflammatory sequelae which complicate recovery. ETT design and materials used have yet to adapt to address intubation associated complications. In this study, a composite coating of electrospun polycaprolactone (PCL) fibers embedded in a four-arm polyethylene glycol acrylate matrix (4APEGA) is developed to transform the ETT from a mechanical device to a dual-purpose device capable of delivering multiple therapeutics while preserving coating integrity. Further, the composite coating system (PCL-4APEGA) is capable of sustained delivery of dexamethasone from the PCL phase and small interfering RNA (siRNA) containing polyplexes from the 4APEGA phase. The siRNA is released rapidly and targets smad3 for immediate reduction in pro-fibrotic transforming growth factor-beta 1 (TGFϐ1) signaling in the upper airway mucosa as well as suppressing long-term sequelae in inflammation from prolonged intubation. A bioreactor was used to study mucosal adhesion to the composite PCL-4APEGA coated ETTs and investigate continued mucus secretory function in ex vivo epithelial samples. The addition of the 4APEGA coating and siRNA delivery to the dexamethasone delivery was then evaluated in a swine model of intubation injury and observed to restore mechanical function of the vocal folds and maintain epithelial thickness when observed over 14 days of intubation. This study demonstrated that increase in surface lubrication paired with surface stiffness reduction significantly decreased fibrotic behavior while reducing epithelial adhesion and abrasion.Biomedical Engineering and Chemical Engineerin
Practical Parallel Algorithms Over GIS Polygonal Datasets for Segment Tree-Based Clipping and for Quad Tree-Based Encoding to Search for Similar Shapes
No full textThe full text of this item is not available at this time because the author has placed this item under an embargo until September 11, 2026.Polygons are fundamental data structures in Geographic information systems (GIS), computer graphics, and CAD. In GIS, polygons represent the boundaries of regions or objects on maps. This dissertation focuses on two critical GIS geometric operations: polygon clipping which is useful in GIS polygon overlay and shape-based similarity search which is useful in spatial data mining and spatial databases. Polygon clipping refers to the computationally expensive task of calculating the common region between two input polygons. While GPU clipping algorithms outperform their CPU multicore counterparts while handling large GIS datasets, they are typically limited to all-to-all-edge intersection tests, and they do not handle degenerate cases properly. Our solution presents a CREW PRAM clipping algorithm and parallelization of Foster’s GPU algorithm handling degenerate cases properly. In real-world polygon clipping, most edge pairs do not contribute to an intersection, which is not addressed by the current practical parallel clipping algorithms. We address this key limitation by eliminating the non-intersecting edge pairs prior to expensive intersection calculations employing filtering techniques and specialized data structures. In our first work, we employ three GPU filters: a common minimum bounding rectangle filter (CMBR), a Count-based filter (CF), and a Line segment minimum bounding rectangle filter (LSMBR). These filters can eliminate a substantial number of non-intersecting edge pairs before performing expensive intersection calculations. Next, we present the first work to extend the segment tree data structure for polygon clipping leveraging Chaselle’s PRAM augmentation, which further enhances non-intersecting edge pair filtering, reducing required work from O(n^2) to Ω(n log n) where n is the number of vertices. Additionally, we demonstrate that it can handle one-to-many polygon clipping efficiently where multiple polygons are clipped against a based polygon. Polygon clipping is challenging in some GIS polygonal datasets since the polygons can change their location and shape over time. Such instances require the identification of polygons between time frames before clipping. Shape-based similarity search over GIS data refers to scanning a reference dataset to find the most similar polygons or contours to a given query shape. In the literature, we do not find scalable shape-based similarity search systems over GIS polygonal data for fast query search. We address this gap by proposing a scalable system utilizing the Jaccard distance metric and a Hierarchical Navigable Small World (HNSW) index over encoded GIS polygons. We present a novel GIS polygon encoding technique leveraging Quad trees to address the challenges posed by the exponential area variability of real-world datasets.Computer Scienc
Episode 2: The role of Tier One museums and libraries
No full textHow do great museums and libraries support great cities and research universities? Explore this and other topics with Dean Hendrix, vice provost of the UTSA Libraries and University Librarian. Listen as Hendrix describes the relationship between the ITC and UTSA, shares the many opportunities that academic libraries and museums offer their communities, and imagines the potential of the ITC to reach learners of all ages in Texas and beyond.Institute of Texan Culture
Evaluating State-of-the-Art CNN Architectures for Multi-Class Casting Defect Detection in Smart Manufacturing
No full textCasting defect detection is crucial in modern manufacturing, particularly in industries such as automotive, aerospace, and heavy machinery. While traditional manual inspection methods remain common, they are time-consuming, labor-intensive, and prone to human error. This study addresses the growing need for automated inspection systems in Industry 4.0 by comparing the performance of four state-of-the-art convolutional neural networks (CNNs)?MobileNetV2, ResNet50, Xception, and AlexNet?for casting defect detection. Using a dataset of 7,348 images from Kaggle, defects were manually categorized into five distinct classes, enabling a more granular analysis than previous binary classification approaches. This comparative evaluation reveals that ResNet50 and Xception demonstrate superior performance, achieving accuracies of 95.22% and 95.71% respectively, with ResNet50 showing exceptional recall (99.92%) and precision (98.51%). MobileNetV2 offers competitive performance (95.01% accuracy) while maintaining computational efficiency, making it suitable for real-time applications. This study provides valuable insights into implementing automated defect detection systems in smart manufacturing environments, particularly where specific defect type identification is crucial for quality control.Mechanical Engineerin
Identification of Risk Factors Linked to Diabetic Foot Ulcers From Multimodal Datasets Using Machine Learning
Get PDFDiabetic Foot Ulcer (DFU) represents a significant and severe complication of diabetes mellitus (DM), frequently resulting in hospitalization and non-traumatic amputations, particularly among Hispanic populations in the United States, where the prevalence rate exceeds 10%. This study employs a multimodal approach, integrating Electronic Health Records (EHRs), bulk RNA sequencing (RNA-seq), and single-cell RNA sequencing (scRNA-seq) to identify and characterize the clinical, molecular, and cellular risk factors associated with DFUs with a specific focus on Hispanic individuals with diabetes.
Machine learning techniques were applied to analyze a wide array of laboratory tests documented in EHRs, including those classified under Logical Observation Identifiers Names and Codes (LOINC). The Albumin/Creatinine Ratio (ACR) test emerged as a key predictor of DFU risk, particularly highlighting its differential impact across ethnic groups. Additionally, the study identified significant gene expression profiles linked to DFU susceptibility and severity, with the Apolipoprotein E (APOE) gene playing a crucial role in non-healing DFUs.
The scRNA-seq analysis provided insights into the cellular heterogeneity within DFU tissues, revealing distinct cell populations, such as keratinocytes, epithelial cells, and macrophages, that contribute to these ulcers' chronic and non-healing nature. The inclusion of Spatial Transcriptomics further enriched the analysis by mapping gene expression patterns to specific tissue locations, offering a more comprehensive understanding of the spatial organization of cells within DFU tissues and their impact on disease progression.
This research demonstrates the potential of integrating multi-omics data with advanced machine learning to enhance the prediction, diagnosis, and treatment of DFUs, particularly in high-risk populations like Hispanics. The findings underscore the importance of personalized medicine approaches in improving healthcare outcomes for diverse population groups, paving the way for more targeted and effective interventions in the management of diabetic complications.Electrical and Computer Engineerin
It Takes a Village: How Community-Based Peer Support for Breastfeeding Bolsters Lactation Prevalence Among Black Mississippians on the Gulf Coast
Get PDFBackground/Objectives: Breastfeeding rates are considerably lower among African American women and across the U.S. South. Our study introduces the concept of <i>community-based peer support for breastfeeding</i>, as measured through beliefs about women&rsquo;s comfort breastfeeding in various social situations (i.e., in the presence of women and men as well as close friends and strangers). Methods: We examine if community-based peer support for breastfeeding is associated with reported lactation prevalence in primary social networks among survey respondents living on the Mississippi Gulf Coast. Special attention is paid to racial differences in the breastfeeding support&ndash;prevalence relationship. We use data drawn from a survey that combines a random sample of adults who are representative of the Mississippi Gulf Coast population and a non-random oversample of African Americans in this predominantly rural tri-county area. Results: Analyses of data from wave 1 of the CDC-funded 2019 Mississippi REACH Social Climate Survey reveal low overall levels of African American breastfeeding network prevalence (knowing friends and family who have breastfed). However, community-based peer support for breastfeeding significantly amplifies breastfeeding network prevalence for black Mississippians when compared with their white counterparts. Discussion: Previous research has indicated that breastfeeding promotional messages have a limited impact on African American breastfeeding propensity along the Mississippi Gulf Coast. However, the current study indicates that enhanced community-based peer support for breastfeeding can be a key facilitator for improved lactation outcomes among African Americans as compared with whites. Conclusion: We establish that breastfeeding is best conceived as both an interpersonal encounter (an activity often conducted in the presence of others) and a collective achievement (a practice influenced by community norms). We discuss study implications and directions for future research.Sociology and Demograph