Multi-Modal Understanding and Generation for Medical Images and Text via Vision-Language Pre-Training

• Published in: IEEE journal of biomedical and health informatics Vol. 26; no. 12; pp. 6070 - 6080
• Main Authors: Moon, Jong Hak, Lee, Hyungyung, Shin, Woncheol, Kim, Young-Hak, Choi, Edward
• Format: Journal Article
• Language: English
• Published: United States IEEE 01.12.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Cognitive tasks; Feature extraction; Healthcare; Humans; Image classification; Language; medical; Medical diagnosis; Medical diagnostic imaging; Medical imaging; Medical Records; MIMICs; multimodal learning; Questions; Radiology; representation learning; self-supervised learning; Task analysis; Training; Vision; vision-and-language; Visualization; X-rays
• ISSN: 2168-2194; 2168-2208
• DOI: 10.1109/JBHI.2022.3207502

Recently a number of studies demonstrated impressive performance on diverse vision-language multi-modal tasks such as image captioning and visual question answering by extending the BERT architecture with multi-modal pre-training objectives. In this work we explore a broad set of multi-modal representation learning tasks in the medical domain, specifically using radiology images and the unstructured report. We propose Medical Vision Language Learner (MedViLL), which adopts a BERT-based architecture combined with a novel multi-modal attention masking scheme to maximize generalization performance for both vision-language understanding tasks (diagnosis classification, medical image-report retrieval, medical visual question answering) and vision-language generation task (radiology report generation). By statistically and rigorously evaluating the proposed model on four downstream tasks with three radiographic image-report datasets (MIMIC-CXR, Open-I, and VQA-RAD), we empirically demonstrate the superior downstream task performance of MedViLL against various baselines, including task-specific architectures.