VALD-MD: Visual Attribution via Latent Diffusion for Medical Diagnostics

TL;DR

This paper introduces VALD-MD, a novel generative approach using latent diffusion models and language prompts to generate normal medical images, enabling visual attribution and interpretability for diagnostics.

Contribution

It presents a new method combining latent diffusion and language models to generate normal counterparts of abnormal images for better interpretability in medical diagnostics.

Findings

Quantitative evaluation with FID, SSIM, MS-SSIM metrics shows high-quality image generation.

System demonstrates zero-shot localized disease induction capabilities.

Effective in highlighting diagnostically relevant image components.

Abstract

Visual attribution in medical imaging seeks to make evident the diagnostically-relevant components of a medical image, in contrast to the more common detection of diseased tissue deployed in standard machine vision pipelines (which are less straightforwardly interpretable/explainable to clinicians). We here present a novel generative visual attribution technique, one that leverages latent diffusion models in combination with domain-specific large language models, in order to generate normal counterparts of abnormal images. The discrepancy between the two hence gives rise to a mapping indicating the diagnostically-relevant image components. To achieve this, we deploy image priors in conjunction with appropriate conditioning mechanisms in order to control the image generative process, including natural language text prompts acquired from medical science and applied radiology. We perform experiments and quantitatively evaluate our results on the COVID-19 Radiography Database containing labelled chest X-rays with differing pathologies via the Frechet Inception Distance (FID), Structural Similarity (SSIM) and Multi Scale Structural Similarity Metric (MS-SSIM) metrics obtained between real and generated images. The resulting system also exhibits a range of latent capabilities including zero-shot localized disease induction, which are evaluated with real examples from the cheXpert dataset.