Med-UniC: Unifying Cross-Lingual Medical Vision-Language Pre-Training by Diminishing Bias

TL;DR

Med-UniC introduces a novel cross-lingual medical vision-language pre-training framework that reduces community bias by aligning semantic representations across English and Spanish medical reports, improving performance across multiple tasks.

Contribution

The paper proposes Cross-lingual Text Alignment Regularization (CTR) with latent language disentanglement to unify cross-lingual medical representations without relying on negative samples.

Findings

Med-UniC outperforms existing methods on 5 medical image tasks.

Reducing community bias improves performance in vision-language and uni-modal visual tasks.

The framework effectively unifies multi-modal data across diverse linguistic communities.

Abstract

The scarcity of data presents a critical obstacle to the efficacy of medical visionlanguage pre-training (VLP). A potential solution lies in the combination of datasets from various language communities. Nevertheless, the main challenge stems from the complexity of integrating diverse syntax and semantics, language-specific medical terminology, and culture-specific implicit knowledge. Therefore, one crucial aspect to consider is the presence of community bias caused by different languages. This paper presents a novel framework named Unifying Cross-Lingual Medical Vision-Language Pre-Training (Med-UniC), designed to integrate multimodal medical data from the two most prevalent languages, English and Spanish. Specifically, we propose Cross-lingual Text Alignment Regularization (CTR) to explicitly unify cross-lingual semantic representations of medical reports originating from diverse language communities. CTR is optimized through latent language disentanglement, rendering our optimization objective to not depend on negative samples, thereby significantly mitigating the bias from determining positive-negative sample pairs within analogous medical reports. Furthermore, it ensures that the cross-lingual representation is not biased toward any specific language community. Med-UniC reaches superior performance across 5 medical image tasks and 10 datasets encompassing over 30 diseases, offering a versatile framework for unifying multi-modal medical data within diverse linguistic communities. The experimental outcomes highlight the presence of community bias in cross-lingual VLP. Reducing this bias enhances the performance not only in vision-language tasks but also in uni-modal visual tasks.