Unsupervised Multimodal Representation Learning across Medical Images and Reports
This work addresses multimodal representation learning for medical applications, but it is incremental as it focuses on establishing baselines rather than introducing major innovations.
The paper tackled the problem of learning joint embeddings between medical images and radiology reports, establishing baseline results on the MIMIC-CXR dataset and showing that limited supervision yields retrieval performance comparable to fully-supervised methods.
Joint embeddings between medical imaging modalities and associated radiology reports have the potential to offer significant benefits to the clinical community, ranging from cross-domain retrieval to conditional generation of reports to the broader goals of multimodal representation learning. In this work, we establish baseline joint embedding results measured via both local and global retrieval methods on the soon to be released MIMIC-CXR dataset consisting of both chest X-ray images and the associated radiology reports. We examine both supervised and unsupervised methods on this task and show that for document retrieval tasks with the learned representations, only a limited amount of supervision is needed to yield results comparable to those of fully-supervised methods.