MedCodER: A Generative AI Assistant for Medical Coding

TL;DR

MedCodER is a novel Generative AI framework that automates medical coding by integrating extraction, retrieval, and re-ranking, achieving high accuracy and providing supporting evidence for code predictions.

Contribution

This work introduces MedCodER, a new generative AI model for medical coding that outperforms existing methods and includes a new annotated dataset with evidence texts.

Findings

Achieves a micro-F1 score of 0.60 on ICD code prediction.

Outperforms state-of-the-art methods in medical coding.

Performance depends on the integration of extraction, retrieval, and re-ranking components.

Abstract

Medical coding is essential for standardizing clinical data and communication but is often time-consuming and prone to errors. Traditional Natural Language Processing (NLP) methods struggle with automating coding due to the large label space, lengthy text inputs, and the absence of supporting evidence annotations that justify code selection. Recent advancements in Generative Artificial Intelligence (AI) offer promising solutions to these challenges. In this work, we introduce MedCodER, a Generative AI framework for automatic medical coding that leverages extraction, retrieval, and re-ranking techniques as core components. MedCodER achieves a micro-F1 score of 0.60 on International Classification of Diseases (ICD) code prediction, significantly outperforming state-of-the-art methods. Additionally, we present a new dataset containing medical records annotated with disease diagnoses, ICD codes, and supporting evidence texts (https://doi.org/10.5281/zenodo.13308316). Ablation tests confirm that MedCodER's performance depends on the integration of each of its aforementioned components, as performance declines when these components are evaluated in isolation.