DDx_CaseReports: LLM Differential Diagnosis Evaluation
What it is
This repository contains code and datasets for a study evaluating how laboratory test results impact the accuracy of differential diagnoses generated by large language models. It includes Jupyter notebooks and scripts to benchmark models like GPT-4 and Llama-2 on 50 clinical case vignettes derived from PubMed Central.
The problem it solves
Uncertainty regarding the specific contribution of laboratory test results to the accuracy of LLM-generated differential diagnoses in clinical settings.
What you could use it for
- Benchmarking LLM performance in clinical differential diagnosis
- Analyzing the specific impact of lab test data on diagnostic accuracy
- Reproducing research on LLMs in medical decision support
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.