Description-based Label Attention Classifier for Explainable ICD-9 Classification

TL;DR

This paper introduces a description-based label attention classifier that enhances explainability in ICD-9 coding from clinical notes, leveraging transformer encoders to improve accuracy and interpretability in noisy medical texts.

Contribution

It proposes a novel label attention mechanism based on label descriptions, improving explainability and performance in ICD-9 classification over existing CNN and RNN models.

Findings

Achieves strong classification results on MIMIC-III-50 dataset.

Provides augmented explainability for clinical note coding.

Outperforms traditional CNN and RNN approaches.

Abstract

ICD-9 coding is a relevant clinical billing task, where unstructured texts with information about a patient's diagnosis and treatments are annotated with multiple ICD-9 codes. Automated ICD-9 coding is an active research field, where CNN- and RNN-based model architectures represent the state-of-the-art approaches. In this work, we propose a description-based label attention classifier to improve the model explainability when dealing with noisy texts like clinical notes. We evaluate our proposed method with different transformer-based encoders on the MIMIC-III-50 dataset. Our method achieves strong results together with augmented explainablilty.