Optimizing Automated KCD Coding: A Retrieval-Verification Approach.

Journal: Studies in health technology and informatics
Published Date: May 15, 2025

Abstract

This study proposes a two-step Retrieval-Verification system for automating the assignment of Korean Standard Classification of Diseases (KCD) codes to free-text diagnoses. The system uses SapBERT-XLMR for initial retrieval, followed by Llama 3.1 for final verification and code selection. Combining the two models improved accuracy to 82.3%. Future work aims to improve the system's performance on abbreviations and conduct experiment with a larger dataset.

Authors

  • Sangji Lee
    Department of Digital Health, Samsung Advanced Institute for Health Science & Technology (SAIHST), Sungkyunkwan University, Seoul, Republic of Korea.
  • Won Chul Cha
    Department of Emergency Medicine, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea.

Keywords

External Resources

Popular Topics
Recent Journals