Dec 1, 2024Press Release

ARISE Awarded $200,000 Stanford Bio-X Seed Grant to Advance Digital Specialty Care with AI

Ethan Goh
ARISE Awarded $200,000 Stanford Bio-X Seed Grant to Advance Digital Specialty Care with AI

The ARISE research network has been awarded a $200,000 Interdisciplinary Initiatives Program Seed Grant for the project “Enhancing Specialty Care with Digital Medical Consultations: A Retrieval-Augmented Language Model Approach.”

The project is led by Dr. Jonathan H. Chen (Biomedical Informatics Research) with collaborators Dr. Michael Bernstein (Computer Science), Dr. Robert Tibshirani (Biomedical Data Science and Statistics), and Dr. Mary Kane Goldstein (Health Policy). Dr. Fateme Nateghi, Postdoctoral Scholar at Stanford, will contribute expertise in applying machine learning and large language model methods to electronic health records, with a focus on evaluation to ensure accuracy.

Addressing Gaps in Specialty Care

Over 25 million people in the United States lack adequate access to medical specialty expertise, disproportionately affecting marginalized populations. While current electronic consultation (eConsult) systems improve access, they remain constrained by specialist labor shortages.

This project will develop and evaluate a Retrieval-Augmented Language Model (RALM) that integrates large language models (LLMs) with patient-specific medical records and evidence-based references. The system will predict diagnostics and medications a specialist might order for a consultation and generate explanatory notes—potentially streamlining access, reducing delays, and preventing unnecessary in-person visits.

Research Plan

Evaluation will focus on:

  • Comparing AI-predicted orders with actual physician decisions
  • Assessing explanatory consultation notes for accuracy and clarity
  • Testing usability and integration into clinical workflows

Why It Matters

This project aims to demonstrate that digital consultation tools can expand healthcare capacity, improve equity, and deliver timely specialty-level guidance to underserved patients.

Findings will create a foundation for broader implementation of AI-supported care delivery.

Full annoucement: https://biox.stanford.edu/research/seed-grants/interdisciplinary-initiatives-program-seed-grant-enhancing-specialty-care