Paying Per-label Attention for Multi-label Extraction from Radiology Reports

TL;DR

This paper introduces a deep learning approach with label-dependent attention to automatically extract structured labels from radiology reports, reducing the need for manual annotation in medical image analysis.

Contribution

It proposes a novel label-dependent attention mechanism combined with synthetic data augmentation for multi-label extraction from radiology reports.

Findings

Robust extraction of multiple labels with a single model

Effective classification of labels as positive, uncertain, or negative

Improved label detection accuracy in radiology reports

Abstract

Training medical image analysis models requires large amounts of expertly annotated data which is time-consuming and expensive to obtain. Images are often accompanied by free-text radiology reports which are a rich source of information. In this paper, we tackle the automated extraction of structured labels from head CT reports for imaging of suspected stroke patients, using deep learning. Firstly, we propose a set of 31 labels which correspond to radiographic findings (e.g. hyperdensity) and clinical impressions (e.g. haemorrhage) related to neurological abnormalities. Secondly, inspired by previous work, we extend existing state-of-the-art neural network models with a label-dependent attention mechanism. Using this mechanism and simple synthetic data augmentation, we are able to robustly extract many labels with a single model, classified according to the radiologist's reporting (positive, uncertain, negative). This approach can be used in further research to effectively extract many labels from medical text.