Skip to content
Emergency Medicine ResearchUniversity of Virginia

Emergency Medicine ResearchUniversity of Virginia

Informatics, AI & data science

The data, models and methods that run through every stage.

What we ask Lens Geriatrics
Data moving through the patient’s journey into a modelSeven stages, from home and community to global, each add a strand of data, monitor waveforms or record entries, to one bundle that runs along the journey into a model. The model’s risk signal stays low, then rises past a threshold, and one risk is flagged.ModelRisk signalRisk flagged

03Emergency department · Arrival & triage

AI flags emergency patients at risk of undiagnosed dementia.

Read the source
IllustrationEvery stage of the journey feeds data into models that can flag risk early. Choose a stage for one example of our work there.

The problem

The problem

Informatics, AI and data science run through every stage of the journey. Our data scientists engineer pipelines for physiological monitor data and health records; we build and evaluate prediction models on emergency department data, model how patients move through the department, weigh which tests and pathways are worth their cost, and study how to put these tools into practice safely.

Two patients’ bedside monitors and a nurse’s tablet streaming data to one dataset board, read by a data scientist at a desk

Focus areas

What this work wants to know

  1. Which patients are at risk of a condition no one has diagnosed yet?

  2. How should prediction models be validated before they reach the bedside?

  3. Where do patients wait, and what would shorten it?

  4. Which tests change management, and which only add time and cost?

  5. How do we turn raw monitor and record data into research-ready datasets?

Studies, papers & news

Featured work

People

Who works on this

Faculty whose research connects to this area.

Meet everyone

  • R. Andrew Taylor, MD, MHS

    Robert E. O’Connor Distinguished Professor & Vice Chair for Research and Innovation

    Research interests: Artificial Intelligence · Clinical Decision Support · Informatics · Digital Health

  • Moira E. Smith, MD, MPH

    Assistant Professor of Emergency Medicine

    Research interests: Clinical Informatics · Clinical Decision Support · AI in Emergency Medicine · Digital Health

  • Rupesh Silwal, PhD

    Data Scientist

    Research interests: Machine Learning · Healthcare Analytics · Large Language Models · Clinical Risk Prediction

  • Jonathan Swap, MSDS

    Data Scientist

    Research interests: Machine Learning · Anomaly Detection · Data Analysis

  • Pavel Chernyavskiy, PhD

    Assistant Professor of Biostatistics

    Research interests: Spatial Statistics · Bayesian Methods · Hierarchical Models · Population Health

  • Thomas R. Hartka, MD, MS, MSDS

    Associate Professor & EMRO Director

    Research interests: Injury Prediction · Crash Biomechanics · Data Science · Prehospital Triage

    In this area

    Dr. Thomas Hartka and UVA colleagues secure third place in the DARPA Triage Challenge

    UVA Emergency Medicine · 2025

  • George F. Glass III, MD

    Associate Professor & Assistant Research Director, EMRO

    Research interests: Operations · Emergency Management · Predictive Analytics · Trauma Biomechanics

  • Andrew E. Muck, MD, MBA

    Professor & Chair, Marcus L. Martin Distinguished Professor

    Research interests: ED Throughput & AI · Medical Direction · Global Health · Medical Education

Get involved

Work with this team

Bring a question, a technology to test, or yourself as a participant.

  • Contact the team

    Questions about this research, the people behind it, or working together.

  • Train with us

    Postdocs, PhD students, fellows and research-minded residents: how to join this work.

  • Interested in partnering on this work?

    Cavalier Catalyst connects industry with emergency physicians, engineers and data scientists.

  • Join a study

    What research in the emergency department involves, how consent works, and who to ask.