Multi-Label Classification of Patient Notes a Case Study on ICD Code Assignment

TL;DR

This paper explores multi-label classification of patient notes for ICD code assignment, introducing a hierarchical attention-based model that improves accuracy and interpretability in electronic health records.

Contribution

The paper presents HA-GRU, a novel hierarchical attention model that achieves state-of-the-art performance and enhances interpretability in ICD code assignment from patient notes.

Findings

HA-GRU outperforms existing models on MIMIC datasets.

Sentence-level attention aids in error analysis and model transparency.

Model provides insights into relevant text segments for each ICD code.

Abstract

In the context of the Electronic Health Record, automated diagnosis coding of patient notes is a useful task, but a challenging one due to the large number of codes and the length of patient notes. We investigate four models for assigning multiple ICD codes to discharge summaries taken from both MIMIC II and III. We present Hierarchical Attention-GRU (HA-GRU), a hierarchical approach to tag a document by identifying the sentences relevant for each label. HA-GRU achieves state-of-the art results. Furthermore, the learned sentence-level attention layer highlights the model decision process, allows easier error analysis, and suggests future directions for improvement.