Reproduce “Conversational Interaction for LLM-Based Clinical Text-to-SQL: A User Study with Healthcare Professionals” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jun-Min Lee, Suyoung Yoo, Gyubok Lee, Edward Choi, Meong Hi Son
The authors evaluate a conversational natural-language-to-SQL system for querying EHRs, which asks users to clarify underspecified requests before generating queries. In a study with 14 physicians using a 100-patient MIMIC-IV demo, mean usability scores rose from 68.2 to 85.9 (p < 0.001). Separately, reviewers marked 134 of 186 conversational turns correct; this was not a comparison of interface accuracy.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.