Document Understanding for Healthcare Referrals

TL;DR

This paper presents a hybrid document understanding model combining LayoutLMv3 and domain-specific rules to improve entity extraction from healthcare referral documents, aiming to reduce costs and errors in patient care workflows.

Contribution

The work introduces a hybrid approach that integrates transformer-based models with domain rules for better accuracy in healthcare referral document analysis.

Findings

Hybrid model significantly improves precision and F1 scores.

Domain-specific rules enhance transformer model performance.

Model effectively handles varying referral document formats.

Abstract

Reliance on scanned documents and fax communication for healthcare referrals leads to high administrative costs and errors that may affect patient care. In this work we propose a hybrid model leveraging LayoutLMv3 along with domain-specific rules to identify key patient, physician, and exam-related entities in faxed referral documents. We explore some of the challenges in applying a document understanding model to referrals, which have formats varying by medical practice, and evaluate model performance using MUC-5 metrics to obtain appropriate metrics for the practical use case. Our analysis shows the addition of domain-specific rules to the transformer model yields greatly increased precision and F1 scores, suggesting a hybrid model trained on a curated dataset can increase efficiency in referral management.