Enhancing LLM Medical Coding with Structured External Knowledge

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Gan, Yidong, Nguyen, David D., Lin, Yang, Zhong, Peter, Vu, Thanh, Duong, Long, Li, Yuan-Fang
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910268653568000
author Gan, Yidong
Nguyen, David D.
Lin, Yang
Zhong, Peter
Vu, Thanh
Duong, Long
Li, Yuan-Fang
author_facet Gan, Yidong
Nguyen, David D.
Lin, Yang
Zhong, Peter
Vu, Thanh
Duong, Long
Li, Yuan-Fang
contents Accurate medical coding requires consulting authoritative resources such as the ICD tabular list and coding guidelines. Existing LLM-based automated methods largely rely on LLMs' internal knowledge, which is prone to hallucination and cannot keep pace with guideline updates. We introduce RAG-Coding, an agentic, training-free method that augments LLMs with structured external knowledge: the tabular list is encoded as a knowledge graph capturing hierarchical and instructional code relationships, and the guidelines are distilled into concise, code-specific summaries rather than retrieved as raw text. To enable our study, we also introduce MDACE-2025, expert re-annotations of the MDACE dataset under the 2025 ICD-10-CM/PCS guidelines, adding code sequencing and justification comments. On MDACE, RAG-Coding outperforms the best LLM-based baseline by 3--13\% in micro-F1 across five LLM backbones, and achieves comparable micro- and macro-F1 to the supervised state-of-the-art, with higher recall ($+$11\%) at the cost of precision ($-$6\%). On MDACE-2025, RAG-Coding outperforms all baselines, demonstrating effective generalisation to updated guidelines. Ablations confirm stepwise gains, highlighting the importance of integrating structured external knowledge for LLM-based medical coding.
format Preprint
id arxiv_https___arxiv_org_abs_2605_27377
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Enhancing LLM Medical Coding with Structured External Knowledge
Gan, Yidong
Nguyen, David D.
Lin, Yang
Zhong, Peter
Vu, Thanh
Duong, Long
Li, Yuan-Fang
Computation and Language
Artificial Intelligence
Information Retrieval
Accurate medical coding requires consulting authoritative resources such as the ICD tabular list and coding guidelines. Existing LLM-based automated methods largely rely on LLMs' internal knowledge, which is prone to hallucination and cannot keep pace with guideline updates. We introduce RAG-Coding, an agentic, training-free method that augments LLMs with structured external knowledge: the tabular list is encoded as a knowledge graph capturing hierarchical and instructional code relationships, and the guidelines are distilled into concise, code-specific summaries rather than retrieved as raw text. To enable our study, we also introduce MDACE-2025, expert re-annotations of the MDACE dataset under the 2025 ICD-10-CM/PCS guidelines, adding code sequencing and justification comments. On MDACE, RAG-Coding outperforms the best LLM-based baseline by 3--13\% in micro-F1 across five LLM backbones, and achieves comparable micro- and macro-F1 to the supervised state-of-the-art, with higher recall ($+$11\%) at the cost of precision ($-$6\%). On MDACE-2025, RAG-Coding outperforms all baselines, demonstrating effective generalisation to updated guidelines. Ablations confirm stepwise gains, highlighting the importance of integrating structured external knowledge for LLM-based medical coding.
title Enhancing LLM Medical Coding with Structured External Knowledge
topic Computation and Language
Artificial Intelligence
Information Retrieval
url https://arxiv.org/abs/2605.27377