University of Houston - Victoria

Graphs are widely used to model relational data in various domains, including social media, e-commerce, telecommunications, and finance. Graph analytics is one of the most popular techniques for analyzing connectivity patterns in communication networks and identifying suspicious behaviors. However, detecting fraud rings at scale is significantly challenging since the volume of graph data is growing exponentially. Moreover, many Python-based graph libraries rely on in-memory computation, which often struggles with large-scale networks. To address these limitations, we extend our previous semiring-based graph computation framework into a domain-specific solution for large-scale fraud ring detection. We formalize key graph patterns that indicate the presence of fraud rings and develop a general detection algorithm based on semiring operations. Our algebraic approach operates efficiently on a hybrid architecture that can scale beyond RAM constraints. Furthermore, it provides interpretable results, ensuring mathematical transparency to meet regulatory demands for explainability. Our approach is developed in C++, and it can be easily called in Python. An Experimental comparison with state-of-the-art Python packages shows that our approach has comparative performance for both small and large graphs.

Bacterial antibiotic resistance has emerged as a significant global threat, making it increasingly challenging to effectively treat infections in patients. Nanomedicine technologies can be implemented for the targeted delivery of medications and drugs to patients. This work investigates the use of silica nanoparticles (SiNPs) loaded with the clindamycin and tetracycline antibiotics to treat Staphylococcus epidermidis infection in a 3D-bioprinted dermal model. SiNPs are stable, biocompatible, and can be loaded with small molecules like antibiotics. The SiNPs were synthesized and loaded with antibiotics. The loading efficiency of the SiNPs was determined by UV-Vis spectroscopy and high-performance liquid chromatography. Dome-shaped constructs containing fibroblasts were 3D printed using a fibrin-based bioink to mimic the dermis of skin. These constructs were then inoculated with bacterial cultures labeled with green fluorescent protein (GFP) 4 days post printing and then treated with antibiotic-loaded SiNPs to determine their effect on bacterial growth. After the incubation phase, the bacteria were cultured in broth to determine the colony-forming unit (CFU) count on toxin superantigen (TSA) plates containing 10 mg/mL chloramphenicol. The CFU count of the 3D-bioprinted human constructs samples treated with antibiotics was significantly lower than both the SiNP-treated and untreated samples. The results suggest that antibioticreleasing SiNPs can serve as a more efficient treatment for skin bacterial infections.

Background and ObjectivesEmergency medical services (EMS) clinicians play a critical role in early identification and management of neurologic emergencies yet may encounter sparse neurology-specific education. With interprofessional collaboration, we developed a multimodal, active-learning, and experiential curriculum for teaching prehospital care in a neurologically underserved setting. We evaluated curriculum feasibility and impact on prehospital neurologic assessment. The objectives of this study were to (1) recognize signs and symptoms of neurologic emergencies, aiding in the formulation of a prehospital differential diagnosis; (2) appraise patient encounter data for incorporation into prehospital medical decision making and facilitation of interprofessional communication and continuity of care; and (3) provide situation-specific prehospital care for patients with neurologic emergencies, including stabilization, medical management, triage, and disposition.MethodsWe delivered a 10-month longitudinal neurologic emergencies curriculum to EMS clinicians in Victoria, TX. The program included case-based learning, simulation, team gamification, and pilot web-based bedside teaching. Participants completed preknowledge and postknowledge assessments for in-person interactive sessions and a simulation, as well as a summative knowledge assessment. Learner reactions were assessed by changes in self-perceived familiarity with neurologic emergencies. We tracked local acute stroke care metrics (monthly median door-to-needle time (DTNT), incidence of thrombolysis, and incidence of thrombectomy transfer) before and after the curriculum to assess impact on recognition and triage skills.ResultsEMS clinicians with varying experience levels participated in 221 session attendances. Short-term knowledge assessment scores improved for neurovascular (60 vs 80, p < 0.01), neuromuscular and spinal cord (60 vs 90, p < 0.01), and traumatic brain injury and headache (60 vs 90, p < 0.01) emergencies, but not for epilepsy/syncope/dizziness/encephalopathy (both 80, p = 0.12), nor summatively (both 80, p = 0.97). Postsimulation assessment showed significant knowledge gain (40 vs 80, p < 0.01). No postcurriculum difference in median DTNT (49 vs 45 minutes, p = 0.72) nor thrombolysis administration (both 2, p = 0.21) was observed; transfer for thrombectomy evaluation increased (1 vs 3, p = 0.02). Learners reported consistently moderate familiarity with neurologic emergencies but expressed appreciation for the case-based, differential diagnosis approach.DiscussionShort-term knowledge improved in several areas of our curriculum, most notably with simulation. Increase in thrombectomy transfers suggests potential impact on recognition and triage. Challenges included attendance amid clinical responsibilities and sample size. Long-term knowledge retention strategies, including ongoing simulation, may help sustain gains.

Abstract In this paper, we propose a novel end‐to‐end method to model 3D objects with geometric details from a single‐view sketch input. Specifically, a novel deep learning‐based differentiable sketch renderer is introduced to establish the relationship between geometric features, represented by normal maps, and 2D sketch strokes. Then, building upon this renderer, we design algorithms to automatically create 3D models with geometric details from a single‐view sketch. With the aid of two introduced loss functions: one based on silhouette‐derived confidence maps and the other on regression similarities, our framework supports the gradient of loss functions calculated between the rendered sketch and input sketch back‐propagating through the whole architecture, thereby enhancing the geometric details on the frontal surface of the generated 3D object. Through comparisons with state‐of‐the‐art sketch‐based 3D modelling techniques, our approach demonstrates superior capability in generating plausible geometric shapes and details, without the necessity for semantic annotations within the input sketch.

Geosynthetic products are widely used in construction projects. Although many studies have demonstrated that geosynthetic solutions result in lower carbon dioxide emissions compared to traditional methods, precise embodied carbon (EC) values for geosynthetic products are scarce. In addition, the EC values of geosynthetics are sometimes substituted with primary raw material data in project carbon footprint calculations, which undermines the credibility of their sustainability claims. This paper reviews EC calculation methods for geosynthetics and provides a geogrid case example. It then extracts EC values from 120 Environmental Product Declarations (EPDs) to propose representative values for geosynthetic products in different regions and recalculates the carbon footprints of geosynthetic-reinforced projects using maximum EC values to assess their impact on project-level estimates. The results show that emissions accumulated during the manufacturing stage of geosynthetic products account for nearly 30% of their total EC. However, the EC of geosynthetic products contributes only a small portion of the total EC of geosynthetic-reinforced projects. Even with higher EC values, geosynthetic solutions remain more sustainable than conventional alternatives. The calculation method presented in this paper enables geosynthetic companies to estimate product EC without commercial life cycle assessment software, while EPD-derived values enhance existing datasets for more accurate carbon footprint calculations in geosynthetic-reinforced projects globally.

Magnetic hyperthermia using heat generation of magnetic nanoparticles (MNPs) has great potential as a noninvasive cancer treatment with minimal side effects. Despite its promising results from basic research, the limited success in clinical translation stems from a lack of systematic nanoparticle optimization. To bridge this gap, machine learning (ML) models that predict the heating efficiency of MNPs composed of different compositions from the physical and magnetic properties of nanoparticles and external magnetic field conditions were developed in this study. The comparative analysis showed that the artificial neural network (ANN), gradient boosting regressor, and random forest regressor outperformed compared to support vector machine (SVM) and k-nearest neighbor (KNN). The ANN models with two hidden layers showed good prediction performances comparative to deeper ANN models. Furthermore, the feature importance analysis revealed the contributions of each input features toward the heating efficiency of MNPs based on the real experimental data.

Health communication campaigns are an effective behavior change tool that can affect a broad range of outcomes, including mental health. During the process of campaign development, a critical decision involves which messages to use in a campaign. One common approach to message selection is the use of target audience message ratings, or what is referred to as perceived message effectiveness (PME). In this approach, target audience members rate messages on PME, and those data are then used to inform message selection for the campaign. The purpose of this chapter is to (a) introduce and define PME, (b) describe the relationship between PME and actual message effectiveness, (c) describe a protocol for applying PME, (d) share two case studies that used PME for message selection, (e) discuss critiques and limitations of PME, and (f) discuss implications for mental health campaign designers.

Abstract This chapter examines the roles of officials who swore oaths to execute the duties of their offices. The transcription of oaths into custumals reflects customary law in action. Oaths were meant to be sworn in public ceremonies, in front of other officials, and frequently in full view of other citizens or burgesses. Customary law in towns was “performed” in public settings through oath-taking, which reinforced the civic identities of the men who held positions of power. This analysis draws from a case study of hundreds of civic oaths and argues that oaths were not only integral to lawmaking in late medieval towns, but also critical to understanding the expectations placed on civic officials. By quantifying and categorizing civic oaths according to their officeholders, as well as the duties and personal qualities these men were expected to embody and perform during their tenures, this chapter shows how anxieties about the morality of civic officials increased over time. This chapter ends with a consideration of the larger implications of civic oath-taking in the wake of the English Reformation, which ushered in large-scale political change.

Background: In 2023, the US state of Texas passed a bill on vaping in school – House Bill 114 (HB 114) – which took effect in Texas public schools on 1 September of that year. HB 114 mandates that any student caught vaping in school be placed on a Disciplinary Alternative Education Program (DAEP). The efficacy of HB 114 may be contingent, at least in part, on the manner in which it is implemented at the school level. This study examined school personnel’s experiences implementing HB 114 in Texas public schools. Methods: We conducted semi-structured interviews with 13 school personnel performing a variety of roles in a Texas public school district. Findings: Thematic analysis of interview data revealed eight themes covering three topics of interest. School personnel reported mixed attitudes towards HB 114 and expressed concern about students adopting more sophisticated covert behaviours to continue vaping undetected. Although vaping prevalence was not measured directly, school personnel did not believe that the policy was reducing the prevalence of student vaping. Schools used diverse strategies to inform students and parents about the policy (e.g. flyers, emails, and social media posting). While some middle and high schools provided vaping prevention education and training for students and school personnel, elementary schools almost exclusively did not have one. The most often cited facilitators for policy implementation were school personnel being cooperative and having the funding to purchase vape detectors. Barriers include limited space and staffing in DAEP, the need for training for school personnel on vaping, and financial strain for more vape detectors. Conclusions: Findings suggest future efforts discouraging vaping should provide additional resources (e.g. financial support) to schools, update school personnel on vaping, offer more prevention and cessation education, and call for more parental and community engagement.