Mamba? Catch The Hype Or Rethink What Really Helps for Image Registration
Bailiang Jian, Jiazhen Pan, Morteza Ghahremani, Daniel Rueckert,, Christian Wachinger, Benedikt Wiestler

TL;DR
This paper argues that complex computational elements do not significantly improve image registration accuracy, emphasizing the value of simple, well-designed methods and rigorous evaluation for better results across various medical imaging tasks.
Contribution
It demonstrates that traditional registration-specific designs outperform complex computational blocks, advocating for simpler solutions and improved evaluation metrics in image registration.
Findings
Advanced computational elements do not significantly improve accuracy.
Well-designed registration-specific methods yield marginal improvements.
Emphasis on rigorous evaluation and simple solutions for diverse imaging modalities.
Abstract
Our findings indicate that adopting "advanced" computational elements fails to significantly improve registration accuracy. Instead, well-established registration-specific designs offer fair improvements, enhancing results by a marginal 1.5\% over the baseline. Our findings emphasize the importance of rigorous, unbiased evaluation and contribution disentanglement of all low- and high-level registration components, rather than simply following the computer vision trends with "more advanced" computational blocks. We advocate for simpler yet effective solutions and novel evaluation metrics that go beyond conventional registration accuracy, warranting further research across diverse organs and modalities. The code is available at \url{https://github.com/BailiangJ/rethink-reg}.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Code & Models
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMedical Image Segmentation Techniques
