Yawen Guo

Hi, I’m Yawen Guo, a Ph.D. student in the Department of Informatics at UC Irvine, advised by Kai Zheng. My research focuses on evaluating healthcare AI systems in real clinical settings and aligning them with human workflows, preferences, and operational constraints.

I combine qualitative inquiry, natural language processing, large language model evaluation, and statistical analytics to study how AI tools reshape clinical documentation, patient-provider communication, data quality, and downstream care outcomes. You can find my publications on Google Scholar and the curated list on my publications page.

current research

• Ambient AI for clinical documentation — evaluating AI-generated notes, clinician edits, documentation burden, specialty-specific customization needs, and the quality/safety implications of human-AI collaboration.
• Patient-provider communication via portals — studying secure messaging patterns, clinician workload, patient communication quality, and AI-assisted triage/drafting tools.
• LLM evaluation for clinical workflows — building practical evaluation pipelines for AI-generated text, including edit/error detection, confidence-aware triage, and evidence-span validation.
• Social determinants of health and predictive modeling — developing SDOH extraction resources and integrating social-risk signals into cardiovascular outcome prediction.
• Population health text analytics — mining social media and video data to understand public attitudes, health literacy, and information quality.

collaborations

• UCI Health
  Clinical AI implementation, ambient documentation evaluation, AI-assisted secure messaging, and workflow QA chatbot pilots for medical trainees.

• Mayo Clinic
  Cardiovascular risk prediction with SDOH, LLM-assisted ontology development, and standardized extraction of social-risk information from clinical notes.

• Keck Medicine of USC
  Predictive analytics for transfer center prioritization, length of stay, readmission, mortality, and structured/free-text drift monitoring.

selected work

• Ambient AI documentation at UCI Health: evaluated real-world AI documentation deployment across hundreds of clinicians, using workflow metrics, surveys, clinician feedback, and large-scale note analyses.
• Clinician edits to AI-drafted notes: developed mixed-methods and LLM-assisted approaches to characterize how clinicians revise ambient AI drafts and where AI systems still require improvement.
• Secure messaging and AI: reviewed patient portal secure messaging research and studied how AI tools can support communication quality while preserving transparency, trust, and equity.
• School-based fitness testing maps: built geospatial tools for long-term California school fitness data to support analysis of health disparities and SDOH.

selected recognition

• Second Place, AMIA Student Paper Competition — 2022
• Best Student Paper Finalist, AMIA Annual Symposium — 2024

teaching and mentorship

• INF 171 — Health Informatics, Teaching Assistant — 2021
• INF 171 — Health Informatics, Guest Lecturer on clinical AI applications — 2024, 2025
• INF 174 — Health Data Analytics, Teaching Assistant — 2026
• Research mentor for undergraduate and master’s students working on secure messaging, ambient AI documentation, school-based fitness testing, and clinical workflow chatbot projects.

service

I have served as a reviewer for Digital Health, Journal of Clinical and Translational Science, npj Health Systems, AMIA Annual Symposium, AMIA Informatics Summit, and IEEE International Conference on Healthcare Informatics. I also volunteered at the AMIA Annual Symposium in 2022.

I build practical, reproducible tools for clinicians and data teams and enjoy collaborative, open science. If you’d like to connect or collaborate, feel free to reach out.