The UF Clinical and Translational Science Institute (CTSI) is pleased to announce the recipients of the 2025–2026 CTS Pilot Award Program (Cycle 1). Eight UF and UF+FSU research teams received funding to develop innovative, reusable tools and methods that advance clinical and translational science — from AI-powered tools that translate complex clinical data into actionable insights, to new approaches for improving patient-provider communication, chronic pain management, and cancer screening. Together, these projects reflect CTSI’s mission to accelerate the translation of scientific discovery into real-world clinical and community impact. Read on to learn more about each project and the difference it aims to make.
Applications for the 2026–2027 CTS Pilot Award Program (Cycle 2) are now open. Investigators interested in pursuing interdisciplinary clinical and translational science research are encouraged to apply. Visit the CTSI Pilot Project Awards page for key dates, eligibility requirements, and submission guidelines.
Pilot Awardees
“Patient-centered communication as teamwork: Development of a simulation-based research tool”
PI: Colleen Gutman, MD
Clinical Associate Professor
In children’s emergency rooms, doctors and nurses often have to communicate quickly with families under stressful, distracting conditions — and one in three families experiences poor communication, which can lead to dangerous medical mistakes. This project develops a simulation-based tool to train doctors and nurses to use teamwork to communicate and partner better with families. With 30 million children receiving ER care annually, the tool will help researchers test and develop new ways to improve communication and, in turn, improve the health of ill and injured children.
“Translating ML-Based HIV Risk Predictions into Clinically Meaningful Narratives Using Large Language Models”
PI: Yiyang Liu, MPH, Ph.D.
Assistant Professor
Artificial intelligence and machine learning tools can help healthcare providers identify patients who may benefit from HIV prevention services, but many current machine learning outputs are difficult for clinicians to understand and trust. This project develops a pipeline that translates complex machine learning-generated HIV risk predictions into clear, clinically meaningful explanations providers can easily interpret — supporting better HIV prevention and earlier identification of patients who may benefit from screening or PrEP. The work could improve identification of at-risk patients while supporting broader, more effective public health implementation of trustworthy AI tools in healthcare.
“Automated Extraction of Immunotherapy-Related Adverse Events and Outcomes”
PI: Protiva Rahman, Ph.D.
Research Assistant Professor
Immune checkpoint inhibitors (ICIs) have greatly advanced cancer treatment and improved survival rates, but they also cause immunotherapy-related adverse events (irAEs) — such as diabetes and fatigue — that affect 30-50% of the patients treated with them, often across melanoma, breast, genitourinary, and lung cancers. Real-world data on irAEs are limited, in part because relevant details are buried in clinical notes. This pilot uses automated, large language model-driven data pipelines to extract irAEs from structured and unstructured UF Health data and build a computable phenotype for endocrine irAEs, providing real-world evidence to improve irAE management and care. The resulting pipeline will be shared through CTSI and the OneFlorida+ NLP workgroup for expansion to other irAE conditions, enabling large-scale, multisite evidence generation.
“Next-Generation Health Promotion: Leveraging Conversational Agent Technology to Expand Motivational Interviewing Access”
PI: Stephanie Staras, PhD
Professor and Associate Chair of Faculty Development
About one-third to one-half of early deaths could be prevented by helping people make healthier choices. Conversations with someone trained in an effective counseling approach can help many people change unhealthy behaviors. However, providing these conversations to everyone would require more healthcare workers and time than are currently available. We propose to use advances in artificial intelligence to make this type of supportive counseling more widely available and help people improve their health and daily habits. We have the tools to help people improve their health and daily habits, and nearly everyone has someone in their life who would benefit from healthier habits. Expanding access to this type of effective counseling approach has potential to dramatically improve health within and outside of the United States.
“Implementation Evaluation of Interoperable Clinical Decision Support for Pain Management as a Reusable Clinical Translational Science Tool“
PI: Khoa Nyguen Pharm.D.
Clinical Associate Professor
Chronic pain is a widespread and costly problem in the U.S., yet doctors often lack practical tools to manage it effectively. Researchers at UF Health are testing two connected digital tools — one for patients to report pain symptoms before a visit, and one that gives doctors a clear summary during the appointment — in real primary care settings across Gainesville and Jacksonville clinics. The goal is a model that can roll out across Florida and beyond, improving pain care while reducing reliance on opioids. This work impacts people living with chronic pain, primary care providers, and healthcare systems, with the potential to transform chronic pain management nationwide and provide clinicians with practical, evidence-based support at the point of care.
“Community Health Data Infrastructure (CHDI) Powered by OneFlorida+: Designing a Reusable Toolkit for Community-Informed Translation of Clinical Data“
PI: Choeeta Chakrabarti, PhD (FSU) — Assistant Professor
MPI: Chris Cogle, MD
Across Florida, hospitals and health systems hold vast amounts of health data that could reveal where disease is concentrated, which communities are being left behind, and what health challenges demand urgent attention — but that data almost never reaches the people who need it most. This project builds a first-of-its-kind toolkit that transforms complex clinical records into clear, privacy-protected health summaries co-designed with community members themselves, ensuring the information produced is built for action, not just academic publication. Health data has long been collected from communities without ever being returned to them; this toolkit breaks that cycle and is designed for any research network in the country to adopt, turning a Florida pilot into a national model for getting health data into the hands of the communities it came from.
“Using Real-World Data to Quantify Cervical Cancer Screening Barriers“
PI: Rahma Mkuu, PhD
Assistant Professor
Cervical cancer is preventable through screening, yet many patients miss recommended screenings due to barriers like discomfort, anxiety, fear, and difficulty accessing care—challenges that often go unnoticed in clinical settings. This project uses artificial intelligence to detect these hidden barriers within doctors’ notes, transforming everyday clinical language into actionable insight. By tapping into data health systems already collect, the goal is to help providers better understand what patients are experiencing and design more personalized support for preventive care. The work has the potential to turn routine clinical documentation into a powerful tool for increasing screening participation—bringing us closer to eliminating cervical cancer altogether.