Learning Semi-Structured Representations of Radiology Reports

TL;DR

This paper introduces a neural sequence-to-sequence model that automatically generates semi-structured representations of radiology reports, improving extraction accuracy and generalization across different datasets.

Contribution

It presents a novel approach combining sentence matching and neural modeling to convert free-text radiology reports into semi-structured formats, enhancing information retrieval.

Findings

Outperforms baseline models on BLEU, ROUGE, METEOR metrics

Achieves qualitative agreement with radiologists

Generalizes well to out-of-sample datasets

Abstract

Beyond their primary diagnostic purpose, radiology reports have been an invaluable source of information in medical research. Given a corpus of radiology reports, researchers are often interested in identifying a subset of reports describing a particular medical finding. Because the space of medical findings in radiology reports is vast and potentially unlimited, recent studies proposed mapping free-text statements in radiology reports to semi-structured strings of terms taken from a limited vocabulary. This paper aims to present an approach for the automatic generation of semi-structured representations of radiology reports. The approach consists of matching sentences from radiology reports to manually created semi-structured representations, followed by learning a sequence-to-sequence neural model that maps matched sentences to their semi-structured representations. We evaluated the proposed approach on the OpenI corpus of manually annotated chest x-ray radiology reports. The results indicate that the proposed approach is superior to several baselines, both in terms of (1) quantitative measures such as BLEU, ROUGE, and METEOR and (2) qualitative judgment of a radiologist. The results also demonstrate that the trained model produces reasonable semi-structured representations on an out-of-sample corpus of chest x-ray radiology reports from a different medical provider.