77710 research outputs found
Sort by
The Kangaroo Care Project: Implementation of Skin-to-Skin (STS) in Sutter Medical Center, Sacramento, Neonatal Intensive Care Unit
No full textBackground: Skin-to-skin contact, also known as kangaroo care or kangaroo mother care, is an intervention typically done after birth to promote physiologic stability, bonding, and breastfeeding.
Aim: The Kangaroo Care project aimed to increase the incidence of skin-to-skin contact in the neonatal intensive care unit at Sutter Medical Center, Sacramento.
Methods: A pre-survey was conducted to assess staff comfort with skin to skin. Interventions to increase skin-to-skin rates included staff e-learning, unplanned extubation mock codes, nurse champions, and a policy change to Sutter Medical Center, Sacramento’s handbook.
Results: TBD. Preliminary results indicated use of STS in the NICU and post-results from the nurses’ survey are pending.
Conclusion: Increasing skin-to-skin care rates in the Sutter Medical Center Sacramento neonatal intensive care unit addressed critical barriers such as staffing shortages, limited space, and equipment availability. By enhancing staff support, education, and training, the NICU fostered greater parental involvement, which is essential for improving outcomes in preterm and low-birth-weight infants. This initiative promoted safer, more effective neonatal care, supported infant growth and development, and strengthened family-centered practices within Sacramento’s diverse community, ultimately contributing to better short- and long-term health outcomes.https://scholarlycommons.pacific.edu/nursing-portfolios/1037/thumbnail.jp
Anti Gambling Gamblers School
No full textAnti Gambling Gamblers School – VR is an immersive educational experience developed for the Meta Quest platform using Unreal Engine. This virtual reality project aims to expose the psychological and mathematical strategies casinos use to manipulate players by placing users inside a lifelike VR casino simulation. Players engage with three popular games—Blackjack, Slots, and Roulette—each designed to mimic real-world casino mechanics while embedding interactive educational layers.
Unlike traditional anti-gambling efforts, this project relies on full VR immersion to make learning experiential and engaging. As users play, in-game prompts and visual overlays explain hidden tactics such as house edge manipulation, near-miss outcomes, and psychological reinforcements like sound design and color schemes. For example, when a user interacts with a virtual slot machine, they can trigger a breakdown of the RNG (Random Number Generator) logic and learn how near-miss outcomes increase addiction risk.
The platform features a Transparency Mode where users can reach out and select elements in the environment—such as “SPIN” buttons or blinking lights—to uncover their hidden purposes (e.g., how MAX BET buttons are designed to encourage impulsive decisions). Embedded pre- and post-experience assessments measure improvements in gambling literacy, including recognition of manipulation tactics, understanding of odds, and changes in behavioral intent.
The VR game will be tested with 50 college students to measure its effectiveness. Metrics tracked include time spent in different game modules, quiz score improvements, and engagement with interactive educational elements. All data is collected securely via Firebase integration.
Developed using Unreal Engine 5 with Meta Quest SDK support, the project showcases real-time 3D environments, gesture-based input, and immersive feedback to simulate authentic casino settings. Deliverables include a complete VR application, an analytical impact report, and a scalable framework for future topics such as loot boxes, sports betting, and mobile gaming.
By combining immersive technology with hands-on education, Anti Gambling Gamblers School – VR empowers users to critically assess the systems designed to exploit them. The goal is not to promote or gamify gambling—but to demystify it. We aim to flip the script and help players understand the game before the game plays them
GateWise: Enhancing Security and Convenience in Gated Communities Through License Plate Recognition
No full textResidential gated communities require efficient vehicle access management while ensuring robust security. Traditional methods often rely on manual processes or outdated technologies that struggle with accuracy, scalability, and adaptability to real-world conditions. To address these challenges, we propose GateWise, an automated vehicle management system leveraging Automated License Plate Recognition (ALPR) technology. GateWise integrates advanced object detection algorithms with a user-friendly web-based interface to streamline vehicle authorization, tracking, and security.
Existing ALPR systems face limitations due to environmental variability (e.g., poor lighting, weather conditions) and dataset constraints, leading to inaccuracies in license plate detection. Open-source models, while cost-effective, often lack diversity in training data, reducing their reliability in dynamic scenarios. These issues compromise both security and user experience, necessitating a more adaptive solution.
GateWise employs open-source deep learning frameworks, specifically YOLOv8, optimized for license plate detection. A centralized web interface enables administrators to add or remove license plates, view all registered plates, and manage permanent or guest access with customizable expiration dates and times, ensuring seamless control over community vehicle permissions.
The development process leverages pretrained YOLOv8 models for both car and license plate detection, enabling efficient and reliable object recognition without the need for custom model training. Detection is tested using a custom-collected dataset that reflects real-world gated community scenarios, including challenging conditions such as glare, rain, and low lighting. Once a license plate is detected, the system uses an OCR engine to extract the alphanumeric text. The extracted text is then refined through post-processing to ensure it complies with California license plate formatting standards, improving the consistency and accuracy of identification. The system is developed using Python-based frameworks, integrating OpenCV for image handling, Tesseract OCR for text extraction, and a lightweight database to manage license plate records.
GateWise is designed to scale across communities of various sizes, with a streamlined, web-based interface that allows administrators to manage vehicle access easily. By focusing on practical, real-world performance and incorporating validation tailored to local standards, GateWise delivers a dependable and adaptable solution for modern vehicle access control in residential communities.
This project demonstrates how AI-driven ALPR systems can balance security and usability. GateWise automates access control while offering administrators granular oversight through its intuitive platform. Future work may explore privacy enhancements and integration with broader smart community ecosystems
DreamDirector: AI-Driven Agentic Platform for Interactive Cinematic Storytelling
No full textDreamDirector is an AI-driven cinematic storytelling platform designed to generate complete interactive narratives from a single user prompt. The goal of the project is to explore how modern generative AI models can be orchestrated together to create a unified, real-time storytelling experience that includes text, images, video, narration, and adaptive music. Instead of requiring users to have expertise in writing, illustration, audio design, or film production, DreamDirector acts as an end-to-end creative engine that automatically produces the components of a visual-novel style story while allowing the user to influence the narrative through meaningful choices.
At the core of the system is a multi-agent architecture built using LangGraph, where different AI agents collaborate to handle narrative writing, visual consistency, media generation, and mood-aware audio. The Story Director agent manages plot progression, branching choices, and the overall structure of the narrative. The Visual Consistency agent maintains coherent character appearances and scene settings across story segments by using reference-based prompting and image-to-image diffusion. The Media Orchestrator coordinates image generation with SDXL, short cinematic video creation using Google Veo, text-to-speech voice narration through ElevenLabs, and real-time adaptive soundscapes generated with Tone.js. Together, these agents enable DreamDirector to deliver a multimedia story experience that updates interactively as the user makes decisions.
The platform is implemented as a full-stack application using React for the front-end, FastAPI for the back-end, and SQLite for persistent story storage. The user interface features a cinematic presentation mode with transitions, subtitles, film grain effects, and a media gallery that organizes all generated assets. Additional tools such as the Casting Director, Location Scout, and Studio Selector extend the platform beyond narrative generation and into early film pre-production workflows by analyzing characters, suggesting real actors, identifying real-world environments, and aligning stories with professional studios.
This project demonstrates the growing potential of AI to support creative work by simulating a collaborative production pipeline across writing, art, audio, and planning. It also highlights practical engineering challenges involving latency, asynchronous media generation, visual consistency, and multimodal synchronization. DreamDirector illustrates how AI can expand access to storytelling and serve as an assistive companion for creators, students, and filmmakers seeking to rapidly prototype narrative ideas
Obsidian Edge
No full textObsidian Edge is a custom-built game engine developed to support rapid creation of 2D card games while deepening familiarity with rendering systems, low-level architecture, and modular engine design. Implemented in C++, the engine features a custom 2D renderer built with GLFW (an open source, multi-platform library for OpenGL development), a high-performance UI system powered by Dear ImGui, audio playback via the Miniaudio library, and fully handmade digital visual assets. The system supports essential card-game operations including, dealing, shuffling, and discarding. This provides a streamlined foundation for prototyping and implementing card-based mechanics.
Its lightweight, modular architecture enables efficient iteration and allows individual subsystems to be extended or replaced with minimal friction. This work also highlights the challenges and design considerations encountered while constructing an engine from the ground up, including rendering performance constraints, asset-management strategies, component organization,and coding for modularity and scalability or for scenarios that could exist. Future development will focus on integrating 3D rendering capabilities, expanding multi-platform support, and introducing scripting through additional programming languages. These improvements aim to increase flexibility and improve usability for larger or more complex projects
Dragon Dash
No full textDragon Dash is a 2D top-down survival and growth game inspired by mechanics from snake.io, developed using Python and the Pygame library. The objective of the game is for the player-controlled dragon to navigate a large scrolling map, collect fish to grow in size, and avoid being caught by enemy dragons. The game features an adaptive camera system that smoothly zooms out as the player grows, providing increased visibility while maintaining gameplay balance. Enemy dragons use a probabilistic AI system that allows them to dynamically choose between chasing the player or pursuing the closest collectible, creating unpredictable and engaging gameplay interactions. The project also implements collision detection, map-boundary constraints, and a resource respawn system to ensure continuous gameplay flow. This game demonstrates fundamental concepts in game design such as movement physics, AI behavior, camera control, and user interaction while maintaining extensibility for future sprite integration and visual enhancements
Smart Honeypot Network with Autonomous Deception
No full textCybersecurity systems traditionally rely on static defensive strategies such as firewalls, intrusion-detection systems, and conventional honeypots that passively record unauthorized access attempts. However, modern cyber threats are increasingly adaptive, automated, and capable of modifying behavior based on the environment encountered. This creates a growing mismatch between dynamic offensive strategies and defensive tools that remain fixed in both structure and response. The problem this project investigates is how a defensive network can engage with an attacker as an intelligence-gathering tool rather than simply blocking intrusion which specifically through interactive deception that adapts to attacker behavior.
This work explores the development of a smart honeypot network capable of autonomous deception. The system is composed of several simulated environments which include SSH, HTTP, and malware-capture honeypots that are deployed within virtual machines to ensure a secure and isolated testing environment. The deception engine was implemented in Python and functions as the central coordinator of adaptive system behavior. It monitors attacker interactions through captured log data and modifies honeypot behavior in real time, such as altering available ports, presenting fabricated file structures, and dynamically changing operating system identity and service banners. This creates the illusion of a responsive live system while remaining entirely artificial.
To evaluate system performance, simulated attacks were conducted using standard penetration tools for reconnaissance, brute-force credential cracking, and controlled exploit and payload testing. All interactions remained confined to a lab environment to ensure ethical and safe experimentation. During these simulations, all commands, timing intervals, session durations, and interaction patterns were logged and analyzed. These logs formed the feature set used for attacker classification.
A key innovation in this project is the integration of an AI-based behavioral classifier. Using extracted features such as command sequencing, probing depth, authentication patterns, and exploration style, the classifier estimates attacker skill level along a defined spectrum. Based on this classification, the deception engine automatically adjusts its response strategy. For example, a less sophisticated attacker may be presented with simpler fake responses, while a more advanced attacker may be guided deeper into fabricated service layers or more convincing system illusions. This results in prolonged engagement and richer intelligence collection.
This project contributes to research in cyber deception and adversarial modeling by demonstrating that a defensive system does not need to merely prevent access as it can actively transform intrusion activity into actionable intelligence. By safely prototyping this approach in a controlled environment, the work lays a foundation for future deployment in real-world network settings, where authentic threat actor behavior can be captured and analyzed with proper safeguards
Bridging oral and systemic health: exploring pathogenesis, biomarkers, and diagnostic innovations in periodontal disease
No full textPurpose This narrative review explores the multifaceted links between periodontal diseases (gingivitis and periodontitis) and systemic health conditions, including cardiovascular disease, diabetes, adverse pregnancy outcomes, Alzheimer’s disease, cancers, rheumatoid arthritis, and respiratory infections. It aims to synthesize evidence on how local oral infections exert systemic effects and evaluate the potential of diagnostic technologies to monitor these interactions. Methods This narrative review synthesizes current scientific literature on periodontal disease pathogenesis, focusing on key pathogens (e.g., Porphyromonas gingivalis, Fusobacterium nucleatum) and their roles in driving local and systemic inflammation via virulence factors and microbial dysbiosis. It examines biomarker-based diagnostic approaches (e.g., IL-1β, TNF-α, microbial DNA) in saliva, blood, and gingival crevicular fluid (GCF) and evaluates current and emerging diagnostic tools (e.g., ELISA, PCR, lateral flow assays, biosensors, microfluidics). Results The review highlights that periodontal pathogens contribute to systemic disease through complex mechanisms including persistent inflammation (driven by cytokines like IL-1β, TNF-α), endotoxemia (via LPS, noting pathogen-specific structural variations impacting immune response), molecular mimicry, and immune modulation. Current diagnostic methods provide valuable information but often face limitations in speed, portability, and multiplexing capability needed for comprehensive point-of-care assessment. Emerging technologies, particularly multiplex platforms integrating biosensors or microfluidics, demonstrate significant potential for rapid, user-friendly analysis of multiple biomarkers, facilitating earlier detection and personalized risk stratification, especially in high-risk populations. Conclusion Periodontal diseases significantly impact systemic health via intricate microbial and inflammatory pathways. The complexity of these interactions necessitates moving beyond conventional diagnostics towards integrated, advanced technologies. Implementing rapid, multiplex biomarker detection platforms within a multidisciplinary healthcare framework holds the potential to revolutionize early detection of linked conditions, improve personalized management strategies, and ultimately reduce the systemic burden of periodontal disease
Circadian rhythmicity and human health
No full textOur modern lifestyle involves irregular eating habits and extended periods spent indoors under artificial lighting. This lifestyle contrasts with the body\u27s circadian rhythms and likely contributes to an increase of chronic diseases worldwide. This special issue on the circadian rhythm contains articles that deepen our understanding of the biological rhythmicity associated with health and disease. Research articles include studies on the effects of light therapy in patients with myocardial infarction and 24-h ambulatory blood pressure monitoring in Japanese women. Review articles cover the roles of micro-RNAs in colorectal cancer, the influence of light, electromagnetic fields and water on biological rhythms, and the effects of eating patterns on metabolic diseases. These studies and review articles highlight the importance of maintaining circadian rhythms and provide practical tips to improve human health
Letter from Henry C. Robinette to Brother, 1863 December 2
No full textHenry Clay Robinette, attended the Delaware Military Academy (1857-1860) and joined the Union Army at the outset of the Civil War. H.C. Robinette fought at the battles of Corinth and Vicksburg (1862) and was later on the General Grant\u27s staff (1864-1865). After the war he was court-martialed for cursing an officer in a barroom brawl (1867)but his father petitioned President Andrew Johnson on his behalf with the result that his sentence was commuted and he was promoted to brevet major for gallant and meritorious services at the Battle of Corinth and the siege of Vicksburg.https://scholarlycommons.pacific.edu/civil-war/1025/thumbnail.jp