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

Registration plays an important role in medical image analysis. Deeplearning-based methods have been studied for medical image registration, whichleverage convolutional neural networks (CNNs) for efficiently regressing adense deformation field from a pair of images. However, CNNs are limited in itsability to extract semantically meaningful intra- and inter-image spatialcorrespondences, which are of importance for accurate image registration. Thisstudy proposes a novel end-to-end deep learning-based framework forunsupervised affine and diffeomorphic deformable registration, referred asACSGRegNet, which integrates a cross-attention module for establishinginter-image feature correspondences and a self-attention module for intra-imageanatomical structures aware. Both attention modules are built on transformerencoders. The output from each attention module is respectively fed to adecoder to generate a velocity field. We further introduce a gated fusionmodule to fuse both velocity fields. The fused velocity field is thenintegrated to a dense deformation field. Extensive experiments are conducted onlumbar spine CT images. Once the model is trained, pairs of unseen lumbarvertebrae can be registered in one shot. Evaluated on 450 pairs of vertebral CTdata, our method achieved an average Dice of 0.963 and an average distanceerror of 0.321mm, which are better than the state-of-the-art (SOTA).

ACSGRegNet: A Deep Learning-based Framework for Unsupervised Joint  Affine and Diffeomorphic Registration of Lumbar Spine CT via Cross- and  Self-Attention Fusion