Generating Radiology Reports via Memory-driven Transformer

TL;DR

This paper introduces a memory-driven Transformer model for automatic radiology report generation, significantly improving report quality and clinical relevance on major datasets, including the first results on MIMIC-CXR.

Contribution

It presents a novel memory-driven Transformer architecture with relational memory and memory-driven normalization for improved medical report generation.

Findings

Outperforms previous models on IU X-Ray and MIMIC-CXR datasets

Able to generate long, detailed reports with medical terms

Produces meaningful image-text attention mappings

Abstract

Medical imaging is frequently used in clinical practice and trials for diagnosis and treatment. Writing imaging reports is time-consuming and can be error-prone for inexperienced radiologists. Therefore, automatically generating radiology reports is highly desired to lighten the workload of radiologists and accordingly promote clinical automation, which is an essential task to apply artificial intelligence to the medical domain. In this paper, we propose to generate radiology reports with memory-driven Transformer, where a relational memory is designed to record key information of the generation process and a memory-driven conditional layer normalization is applied to incorporating the memory into the decoder of Transformer. Experimental results on two prevailing radiology report datasets, IU X-Ray and MIMIC-CXR, show that our proposed approach outperforms previous models with respect to both language generation metrics and clinical evaluations. Particularly, this is the first work reporting the generation results on MIMIC-CXR to the best of our knowledge. Further analyses also demonstrate that our approach is able to generate long reports with necessary medical terms as well as meaningful image-text attention mappings.