BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration

Alan Q. Wang1, Rachit Saluja1, Heejong Kim1, Xinzi He1, Adrian Dalca2, Mert R. Sabuncu3
1: Department of Radiology, Weill Cornell Medicine, New York City, NY, USA, 2: Computer Science and Artificial Intelligence Lab at the Massachusetts Institute of Technology, Cambridge, MA, USA, 3: School of Electrical and Computer Engineering, Cornell University, Ithaca, NY, USA
Publication date: 2025/05/27
https://doi.org/10.59275/j.melba.2025-59g7
PDF · Code

Abstract

We present a keypoint-based foundation model for general purpose brain MRI registration, based on the recently-proposed KeyMorph framework. Our model, called BrainMorph, serves as a tool that supports multi-modal, pairwise, and scalable groupwise registration. BrainMorph is trained on a massive dataset of over 100,000 3D volumes, skull-stripped and non-skull-stripped, from nearly 16,000 unique healthy and diseased subjects. BrainMorph is robust to large misalignments, interpretable via interrogating automatically-extracted keypoints, and enables rapid and controllable generation of many plausible transformations with different alignment types and different degrees of nonlinearity at test-time. We demonstrate the superiority of BrainMorph in solving 3D rigid, affine, and nonlinear registration on a variety of multi-modal brain MRI scans of healthy and diseased subjects, in both the pairwise and groupwise setting. In particular, we show registration accuracy and speeds that surpass many classical and learning-based methods, especially in the context of large initial misalignments and large group settings. All code and models are available at https://github.com/alanqrwang/brainmorph

Keywords

Image registration · Multi-modal · Keypoint detection · Foundation model · Brain MRI

Bibtex @article{melba:2025:010:wang, title = "BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration", author = "Wang, Alan Q. and Saluja, Rachit and Kim, Heejong and He, Xinzi and Dalca, Adrian and Sabuncu, Mert R.", journal = "Machine Learning for Biomedical Imaging", volume = "3", issue = "May 2025 issue", year = "2025", pages = "181--215", issn = "2766-905X", doi = "https://doi.org/10.59275/j.melba.2025-59g7", url = "https://melba-journal.org/2025:010" }
RISTY - JOUR AU - Wang, Alan Q. AU - Saluja, Rachit AU - Kim, Heejong AU - He, Xinzi AU - Dalca, Adrian AU - Sabuncu, Mert R. PY - 2025 TI - BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration T2 - Machine Learning for Biomedical Imaging VL - 3 IS - May 2025 issue SP - 181 EP - 215 SN - 2766-905X DO - https://doi.org/10.59275/j.melba.2025-59g7 UR - https://melba-journal.org/2025:010 ER -

2025:010 cover