Adapting Pretrained Vision-Language Foundational Models to Medical Imaging Domains

TL;DR

This paper adapts large pretrained vision-language models, specifically Stable Diffusion, to generate and manipulate medical images, addressing domain shift issues and improving clinical relevance through fine-tuning and evaluation.

Contribution

It demonstrates how to fine-tune the Stable Diffusion model for medical imaging, enabling realistic abnormality insertion while preserving diagnostic features.

Findings

Improved image quality metrics over baseline models

Radiologist evaluations confirm clinical relevance

Model maintains 95% abnormality detection accuracy

Abstract

Multi-modal foundation models are typically trained on millions of pairs of natural images and text captions, frequently obtained through web-crawling approaches. Although such models depict excellent generative capabilities, they do not typically generalize well to specific domains such as medical images that have fundamentally shifted distributions compared to natural images. Building generative models for medical images that faithfully depict clinical context may help alleviate the paucity of healthcare datasets. Thus, in this study, we seek to research and expand the representational capabilities of large pretrained foundation models to medical concepts, specifically for leveraging the Stable Diffusion model to generate domain specific images found in medical imaging. We explore the sub-components of the Stable Diffusion pipeline (the variational autoencoder, the U-Net and the text-encoder) to fine-tune the model to generate medical images. We benchmark the efficacy of these efforts using quantitative image quality metrics and qualitative radiologist-driven evaluations that accurately represent the clinical content of conditional text prompts. Our best-performing model improves upon the stable diffusion baseline and can be conditioned to insert a realistic-looking abnormality on a synthetic radiology image, while maintaining a 95% accuracy on a classifier trained to detect the abnormality.