Unsupervised Detection of Fetal Brain Anomalies using Denoising Diffusion Models

TL;DR

This paper introduces an unsupervised method using diffusion models for fetal brain anomaly detection in ultrasound images, eliminating the need for abnormal data and addressing limitations of supervised approaches.

Contribution

The work presents a novel inpainting-based diffusion model approach for unsupervised fetal brain anomaly detection using only normal images for training.

Findings

Effective detection of fetal brain anomalies in ultrasound images.

Diffusion models' robustness varies with different noise types.

Potential for clinical application in fetal anomaly screening.

Abstract

Congenital malformations of the brain are among the most common fetal abnormalities that impact fetal development. Previous anomaly detection methods on ultrasound images are based on supervised learning, rely on manual annotations, and risk missing underrepresented categories. In this work, we frame fetal brain anomaly detection as an unsupervised task using diffusion models. To this end, we employ an inpainting-based Noise Agnostic Anomaly Detection approach that identifies the abnormality using diffusion-reconstructed fetal brain images from multiple noise levels. Our approach only requires normal fetal brain ultrasound images for training, addressing the limited availability of abnormal data. Our experiments on a real-world clinical dataset show the potential of using unsupervised methods for fetal brain anomaly detection. Additionally, we comprehensively evaluate how different noise types affect diffusion models in the fetal anomaly detection domain.