An Ensemble Approach for Automatic Structuring of Radiology Reports

TL;DR

This paper introduces an ensemble method combining three models to automatically label sections in radiology reports, addressing variability in style and format, and achieving high accuracy in structuring medical records.

Contribution

The paper presents a novel ensemble approach utilizing Bi-LSTMs and report-specific features for accurate automatic structuring of radiology reports, outperforming existing methods.

Findings

Achieved 97.1% accuracy in report section labeling

Outperformed multiple baseline and state-of-the-art approaches

Validated on proprietary and MIMIC-III datasets

Abstract

Automatic structuring of electronic medical records is of high demand for clinical workflow solutions to facilitate extraction, storage, and querying of patient care information. However, developing a scalable solution is extremely challenging, specifically for radiology reports, as most healthcare institutes use either no template or department/institute specific templates. Moreover, radiologists' reporting style varies from one to another as sentences are telegraphic and do not follow general English grammar rules. We present an ensemble method that consolidates the predictions of three models, capturing various attributes of textual information for automatic labeling of sentences with section labels. These three models are: 1) Focus Sentence model, capturing context of the target sentence; 2) Surrounding Context model, capturing the neighboring context of the target sentence; and finally, 3) Formatting/Layout model, aimed at learning report formatting cues. We utilize Bi-directional LSTMs, followed by sentence encoders, to acquire the context. Furthermore, we define several features that incorporate the structure of reports. We compare our proposed approach against multiple baselines and state-of-the-art approaches on a proprietary dataset as well as 100 manually annotated radiology notes from the MIMIC-III dataset, which we are making publicly available. Our proposed approach significantly outperforms other approaches by achieving 97.1% accuracy.