Spatially-varying Regularization with Conditional Transformer for Unsupervised Image Registration
This work addresses a domain-specific problem in medical imaging by enabling more accurate and adaptable unsupervised image registration, though it is incremental as it builds upon existing Transformer-based models.
The paper tackled the problem of spatially-invariant regularization in deep learning-based image registration by introducing an end-to-end framework that learns a spatially-varying deformation regularizer from data, resulting in significant performance improvement while maintaining smooth deformation.
In the past, optimization-based registration models have used spatially-varying regularization to account for deformation variations in different image regions. However, deep learning-based registration models have mostly relied on spatially-invariant regularization. Here, we introduce an end-to-end framework that uses neural networks to learn a spatially-varying deformation regularizer directly from data. The hyperparameter of the proposed regularizer is conditioned into the network, enabling easy tuning of the regularization strength. The proposed method is built upon a Transformer-based model, but it can be readily adapted to any network architecture. We thoroughly evaluated the proposed approach using publicly available datasets and observed a significant performance improvement while maintaining smooth deformation. The source code of this work will be made available after publication.