Unsupervised Anomaly Detection in Medical Images Using Masked Diffusion Model

TL;DR

This paper introduces a novel unsupervised diffusion-based method with masking techniques for detecting anomalies in medical images, specifically brain MRI scans, without requiring pixel-level labels, achieving superior results over existing methods.

Contribution

It presents masked-DDPM, a new self-supervised diffusion model incorporating Masked Image and Frequency Modeling for improved anomaly detection in medical images.

Findings

Outperforms existing supervised and weakly supervised methods

Effective in detecting tumors and sclerosis lesions

First application of Masked Frequency Modeling in diffusion models for medical imaging

Abstract

It can be challenging to identify brain MRI anomalies using supervised deep-learning techniques due to anatomical heterogeneity and the requirement for pixel-level labeling. Unsupervised anomaly detection approaches provide an alternative solution by relying only on sample-level labels of healthy brains to generate a desired representation to identify abnormalities at the pixel level. Although, generative models are crucial for generating such anatomically consistent representations of healthy brains, accurately generating the intricate anatomy of the human brain remains a challenge. In this study, we present a method called masked-DDPM (mDPPM), which introduces masking-based regularization to reframe the generation task of diffusion models. Specifically, we introduce Masked Image Modeling (MIM) and Masked Frequency Modeling (MFM) in our self-supervised approach that enables models to learn visual representations from unlabeled data. To the best of our knowledge, this is the first attempt to apply MFM in DPPM models for medical applications. We evaluate our approach on datasets containing tumors and numerous sclerosis lesions and exhibit the superior performance of our unsupervised method as compared to the existing fully/weakly supervised baselines. Code is available at https://github.com/hasan1292/mDDPM.