KU-DMIS-MSRA at RadSum23: Pre-trained Vision-Language Model for Radiology Report Summarization

TL;DR

This paper presents CheXOFA, a pre-trained vision-language model tailored for radiology report summarization, which leverages multimodal pre-training and domain adaptation to achieve state-of-the-art results in chest X-ray report summarization.

Contribution

Introduction of CheXOFA, a novel pre-trained vision-language model that adapts general domain multimodal pre-training to the chest X-ray domain for improved summarization.

Findings

Achieved first place on RadSum23 leaderboard.

Demonstrated superior performance on BioNLP shared task datasets.

Effective domain transfer with limited resources.

Abstract

In this paper, we introduce CheXOFA, a new pre-trained vision-language model (VLM) for the chest X-ray domain. Our model is initially pre-trained on various multimodal datasets within the general domain before being transferred to the chest X-ray domain. Following a prominent VLM, we unify various domain-specific tasks into a simple sequence-to-sequence schema. It enables the model to effectively learn the required knowledge and skills from limited resources in the domain. Demonstrating superior performance on the benchmark datasets provided by the BioNLP shared task, our model benefits from its training across multiple tasks and domains. With subtle techniques including ensemble and factual calibration, our system achieves first place on the RadSum23 leaderboard for the hidden test set.