Unifying Heterogeneous Electronic Health Records Systems via Text-Based Code Embedding

TL;DR

This paper introduces DescEmb, a text-based embedding framework for electronic health records that unifies heterogeneous systems and improves transferability across hospitals using neural language models.

Contribution

The paper presents a novel code-agnostic embedding method for EHR that leverages textual descriptions, enabling unified modeling across diverse systems.

Findings

DescEmb outperforms traditional code-based embeddings.

Effective in zero-shot transfer between hospitals.

Supports training a single model for multiple EHR datasets.

Abstract

EHR systems lack a unified code system forrepresenting medical concepts, which acts asa barrier for the deployment of deep learningmodels in large scale to multiple clinics and hos-pitals. To overcome this problem, we introduceDescription-based Embedding,DescEmb, a code-agnostic representation learning framework forEHR. DescEmb takes advantage of the flexibil-ity of neural language understanding models toembed clinical events using their textual descrip-tions rather than directly mapping each event toa dedicated embedding. DescEmb outperformedtraditional code-based embedding in extensiveexperiments, especially in a zero-shot transfertask (one hospital to another), and was able totrain a single unified model for heterogeneousEHR datasets.