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/

Global: Zendy Plus annual

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

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

The Swin transformer has recently attracted attention in medical imageanalysis due to its computational efficiency and long-range modelingcapability. Owing to these properties, the Swin Transformer is suitable forestablishing more distant relationships between corresponding voxels indifferent positions in complex abdominal image registration tasks. However, theregistration models based on transformers combine multiple voxels into a singlesemantic token. This merging process limits the transformers to model andgenerate coarse-grained spatial information. To address this issue, we proposeRecovery Feature Resolution Network (RFRNet), which allows the transformer tocontribute fine-grained spatial information and rich semantic correspondencesto higher resolution levels. Furthermore, shifted window partitioningoperations are inflexible, indicating that they cannot perceive the semanticinformation over uncertain distances and automatically bridge the globalconnections between windows. Therefore, we present a Weighted Window Attention(WWA) to build global interactions between windows automatically. It isimplemented after the regular and cyclic shift window partitioning operationswithin the Swin transformer block. The proposed unsupervised deformable imageregistration model, named RFR-WWANet, detects the long-range correlations, andfacilitates meaningful semantic relevance of anatomical structures. Qualitativeand quantitative results show that RFR-WWANet achieves significant improvementsover the current state-of-the-art methods. Ablation experiments demonstrate theeffectiveness of the RFRNet and WWA designs. Our code is available at\url{https://github.com/MingR-Ma/RFR-WWANet}.

RFR-WWANet: Weighted Window Attention-Based Recovery Feature Resolution Network for Unsupervised Image Registration