New LLM framework KREL enhances automatic medical coding with ICD guidelines

Researchers have developed KREL, a novel framework designed to improve automatic medical coding using large language models (LLMs). KREL integrates external International Classification of Diseases (ICD) coding guidelines as structured knowledge, enabling LLMs to reason over clinical evidence more effectively. This approach aims to address challenges such as the length of clinical notes, the vast ICD label space, and complex coding rules, leading to reduced hallucinations and better compliance with coding standards. Experiments demonstrate that KREL outperforms existing pre-trained language model and LLM-based methods on benchmark datasets. AI

IMPACT This framework could streamline medical billing and improve the accuracy of clinical data analysis by leveraging LLMs for complex coding tasks.

RANK_REASON The item is a research paper detailing a new framework for medical coding. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• alphaXiv
• arXiv
• CatalyzeX
• DagsHub
• Gotit.pub
• Hugging Face
• International Statistical Classification of Diseases and Related Health Problems
• LLMs
• ScienceCast

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →