Rotation Equivariant Convolutions in Deformable Registration of Brain MRI
2026-04-09 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors worked on improving how computers align brain MRI images by making their neural networks understand rotations better. They replaced parts of existing models with special layers that treat rotated images more naturally. This helped their models be more accurate, handle rotated images well, and learn effectively even with less data. Their findings suggest adding knowledge about image geometry helps create better image registration tools.
image registrationbrain MRIrotation equivarianceconvolutional neural networksdeformable registrationgeometric priorssample efficiencyinductive biasencoder networks
Authors
Arghavan Rezvani, Kun Han, Anthony T. Wu, Pooya Khosravi, Xiaohui Xie
Abstract
Image registration is a fundamental task that aligns anatomical structures between images. While CNNs perform well, they lack rotation equivariance - a rotated input does not produce a correspondingly rotated output. This hinders performance by failing to exploit the rotational symmetries inherent in anatomical structures, particularly in brain MRI. In this work, we integrate rotation-equivariant convolutions into deformable brain MRI registration networks. We evaluate this approach by replacing standard encoders with equivariant ones in three baseline architectures, testing on multiple public brain MRI datasets. Our experiments demonstrate that equivariant encoders have three key advantages: 1) They achieve higher registration accuracy while reducing network parameters, confirming the benefit of this anatomical inductive bias. 2) They outperform baselines on rotated input pairs, demonstrating robustness to orientation variations common in clinical practice. 3) They show improved performance with less training data, indicating greater sample efficiency. Our results demonstrate that incorporating geometric priors is a critical step toward building more robust, accurate, and efficient registration models.