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

In 3D medical image segmentation, small targets segmentation is crucial fordiagnosis but still faces challenges. In this paper, we propose the AxisProjection Attention UNet, named APAUNet, for 3D medical image segmentation,especially for small targets. Considering the large proportion of thebackground in the 3D feature space, we introduce a projection strategy toproject the 3D features into three orthogonal 2D planes to capture thecontextual attention from different views. In this way, we can filter out theredundant feature information and mitigate the loss of critical information forsmall lesions in 3D scans. Then we utilize a dimension hybridization strategyto fuse the 3D features with attention from different axes and merge them by aweighted summation to adaptively learn the importance of differentperspectives. Finally, in the APA Decoder, we concatenate both high and lowresolution features in the 2D projection process, thereby obtaining moreprecise multi-scale information, which is vital for small lesion segmentation.Quantitative and qualitative experimental results on two public datasets (BTCVand MSD) demonstrate that our proposed APAUNet outperforms the other methods.Concretely, our APAUNet achieves an average dice score of 87.84 on BTCV, 84.48on MSD-Liver and 69.13 on MSD-Pancreas, and significantly surpass the previousSOTA methods on small targets.

APAUNet: Axis Projection Attention UNet for Small Target in 3D Medical Segmentation