English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Open

Due to their extreme long-range modeling capability, vision transformer-basednetworks have become increasingly popular in deformable image registration. Webelieve, however, that the receptive field of a 5-layer convolutional U-Net issufficient to capture accurate deformations without needing long-rangedependencies. The purpose of this study is therefore to investigate whetherU-Net-based methods are outdated compared to modern transformer-basedapproaches when applied to medical image registration. For this, we propose alarge kernel U-Net (LKU-Net) by embedding a parallel convolutional block to avanilla U-Net in order to enhance the effective receptive field. On the public3D IXI brain dataset for atlas-based registration, we show that the performanceof the vanilla U-Net is already comparable with that of state-of-the-arttransformer-based networks (such as TransMorph), and that the proposed LKU-Netoutperforms TransMorph by using only 1.12% of its parameters and 10.8% of itsmult-adds operations. We further evaluate LKU-Net on a MICCAI Learn2Reg 2021challenge dataset for inter-subject registration, our LKU-Net also outperformsTransMorph on this dataset and ranks first on the public leaderboard as of thesubmission of this work. With only modest modifications to the vanilla U-Net,we show that U-Net can outperform transformer-based architectures oninter-subject and atlas-based 3D medical image registration. Code is availableat https://github.com/xi-jia/LKU-Net.

U-Net vs Transformer: Is U-Net Outdated in Medical Image Registration?