Lybarger Language Lab

About the Lab

We are a natural language processing lab building AI systems that improve how health information is understood, communicated, and exchanged. Our work spans extracting clinical information from medical records, such as social determinants of health, symptoms, and findings from imaging reports; simplifying complex health information for patients; and developing conversational agents for clinical and mental health settings. Across these efforts, we prioritize methods that are trustworthy, equitable, and grounded in real-world healthcare needs.

Prospective PhD students

I expect to admit at least one PhD student to start in fall 2027. Admission runs through the PhD in Information Technology program, not through me, so apply there first.

If you want me to see your application, email klybarge@gmu.edu with "PhD 2027" in the subject line. Introduce yourself, say what you want to work on, and attach your CV. Ideas close to our current work and ideas that point somewhere new are both welcome. I read these in batches and start replying in late fall. Silence before then is about my schedule, not your application.

More about applying to the lab

Projects & Research Areas

Conversational AI for Depression Management Funded

With Dr. Farrokh Alemi, evaluating whether large language models can give trustworthy guidance on antidepressant selection. PCORI-funded. Project details →

Understanding Chronic Pain through Language Funded

Capturing patients' lived experience of chronic knee pain from motivational-interviewing transcripts, with annotation methods and extraction models built for them. NIH-funded. Project details →

Health Literacy & Text Simplification

Making health information easier for patients to understand: annotated corpora for cancer education, readability-controlled question answering, and efficient extraction for health-communication materials. Area details →

NLP for Correctional & Legal Records

Extracting information from correctional and legal case records: probation and parole supervisory notes, and handwritten assault and violence reports. Area details →

  • Federal Probation · 2023

    Exploring Probation and Parole Records Using Natural Language Processing: A Case Study of Supervisory Condition Notes

    Elyazori

  • AMIA Symposium Workshop · 2024 Poster

    Optical Character Recognition for Information Extraction from Handwritten Assault and Violence Reports

    Irbaz et al.

Emerging Directions

Early-stage lines of work, one paper each so far: jailbreak resistance for protected LLMs, and diversity-aware retrieval-augmented generation. Area details →

Explore all projects & research areas

Latest News

Clinical NLP chapter published in the Springer Handbook of Data Engineering

Tanvir's patient simulation paper published in JMIR AI

Flyer for the Healthcare Access panel and roundtable, October 15, 2026

Hadeel to join healthcare access panel at Mason

Tanvir's paper accepted in JMIR AI

All news

Meet the Team

Learn more about the team