Using Artificial Intelligence to Label Free-Text Operative and Ultrasound Reports for Grading Pediatric Appendicitis

In brief

This study evaluates ChatGPT-4's ability to extract structured clinical data from free-text operative and ultrasound reports in pediatric appendicitis cases, comparing AI performance against human data collectors. The work addresses a key barrier to personalized appendicitis management: converting unstructured EMR documentation into analyzable datasets for clinical decision support.

• AI chatbots like ChatGPT-4 can extract structured clinical data from free-text operative and ultrasound reports in pediatric appendicitis cases.
• Small datasets and unstructured EMR data are major barriers to developing personalized appendicitis management algorithms.
• Large language models offer a scalable alternative to manual chart review for converting narrative clinical notes into analyzable datasets.
• Study directly compares AI extraction quality against human data collectors to validate clinical utility of automated approaches.

Written by the GCMD Library team from the article.