Deformable Registration of Brain MR Images via a Hybrid Loss
This work addresses the challenge of improving registration accuracy for medical imaging, specifically for T1-weighted MR images, but it is incremental as it builds on existing unsupervised learning methods with a novel loss design.
The paper tackled the problem of deformable registration of brain MR images by proposing a hybrid loss function that integrates multiple image characteristics, achieving high accuracy on the OASIS dataset while preserving deformation smoothness.
Unsupervised learning strategy is widely adopted by the deformable registration models due to the lack of ground truth of deformation fields. These models typically depend on the intensity-based similarity loss to obtain the learning convergence. Despite the success, such dependence is insufficient. For the deformable registration of mono-modality image, well-aligned two images not only have indistinguishable intensity differences, but also are close in the statistical distribution and the boundary areas. Considering that well-designed loss functions can facilitate a learning model into a desirable convergence, we learn a deformable registration model for T1-weighted MR images by integrating multiple image characteristics via a hybrid loss. Our method registers the OASIS dataset with high accuracy while preserving deformation smoothness.