Denoising Diffusion Models for Anomaly Localization in Medical Images

TL;DR

This paper reviews the use of denoising diffusion models for localizing anomalies in medical images, discussing methodologies, datasets, supervision schemes, challenges, and future research directions.

Contribution

It provides a comprehensive overview of diffusion models applied to medical anomaly localization, highlighting current methods, datasets, evaluation metrics, and open challenges.

Findings

Diffusion models show promise for accurate anomaly localization.

Supervision schemes vary from fully supervised to unsupervised.

Open challenges include detection bias and computational cost.

Abstract

This review explores anomaly localization in medical images using denoising diffusion models. After providing a brief methodological background of these models, including their application to image reconstruction and their conditioning using guidance mechanisms, we provide an overview of available datasets and evaluation metrics suitable for their application to anomaly localization in medical images. In this context, we discuss supervision schemes ranging from fully supervised segmentation to semi-supervised, weakly supervised, self-supervised, and unsupervised methods, and provide insights into the effectiveness and limitations of these approaches. Furthermore, we highlight open challenges in anomaly localization, including detection bias, domain shift, computational cost, and model interpretability. Our goal is to provide an overview of the current state of the art in the field, outline research gaps, and highlight the potential of diffusion models for robust anomaly localization in medical images.