Unsupervised Anomaly Detection in Brain MRI via Disentangled Anatomy Learning

TL;DR

This paper introduces a novel unsupervised brain MRI anomaly detection method that disentangles anatomical information from imaging data and reconstructs pseudo-healthy images, improving generalizability and reducing abnormal residuals across multi-center datasets.

Contribution

The paper proposes a new framework with disentangled representation and edge-to-image restoration modules, enhancing unsupervised anomaly detection in brain MRI by focusing on anatomy and structural details.

Findings

Outperforms 17 state-of-the-art methods in AP and DSC metrics.

Achieves +18.32% in AP and +13.64% in DSC on multi-center datasets.

Demonstrates improved generalizability to multi-modality and multi-center MRIs.

Abstract

Detection of various lesions in brain MRI is clinically critical, but challenging due to the diversity of lesions and variability in imaging conditions. Current unsupervised learning methods detect anomalies mainly through reconstructing abnormal images into pseudo-healthy images (PHIs) by normal samples learning and then analyzing differences between images. However, these unsupervised models face two significant limitations: restricted generalizability to multi-modality and multi-center MRIs due to their reliance on the specific imaging information in normal training data, and constrained performance due to abnormal residuals propagated from input images to reconstructed PHIs. To address these limitations, two novel modules are proposed, forming a new PHI reconstruction framework. Firstly, the disentangled representation module is proposed to improve generalizability by decoupling brain MRI into imaging information and essential imaging-invariant anatomical images, ensuring that the reconstruction focuses on the anatomy. Specifically, brain anatomical priors and a differentiable one-hot encoding operator are introduced to constrain the disentanglement results and enhance the disentanglement stability. Secondly, the edge-to-image restoration module is designed to reconstruct high-quality PHIs by restoring the anatomical representation from the high-frequency edge information of anatomical images, and then recoupling the disentangled imaging information. This module not only suppresses abnormal residuals in PHI by reducing abnormal pixels input through edge-only input, but also effectively reconstructs normal regions using the preserved structural details in the edges. Evaluated on nine public datasets (4,443 patients' MRIs from multiple centers), our method outperforms 17 SOTA methods, achieving absolute improvements of +18.32% in AP and +13.64% in DSC.