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

U-Nets have achieved tremendous success in medical image segmentation.Nevertheless, it may suffer limitations in global (long-range) contextualinteractions and edge-detail preservation. In contrast, Transformer has anexcellent ability to capture long-range dependencies by leveraging theself-attention mechanism into the encoder. Although Transformer was born tomodel the long-range dependency on the extracted feature maps, it still suffersfrom extreme computational and spatial complexities in processinghigh-resolution 3D feature maps. This motivates us to design the efficientlyTransformer-based UNet model and study the feasibility of Transformer-basednetwork architectures for medical image segmentation tasks. To this end, wepropose to self-distill a Transformer-based UNet for medical imagesegmentation, which simultaneously learns global semantic information and localspatial-detailed features. Meanwhile, a local multi-scale fusion block is firstproposed to refine fine-grained details from the skipped connections in theencoder by the main CNN stem through self-distillation, only computed duringtraining and removed at inference with minimal overhead. Extensive experimentson BraTS 2019 and CHAOS datasets show that our MISSU achieves the bestperformance over previous state-of-the-art methods. Code and models areavailable at \url{https://github.com/wangn123/MISSU.git}

MISSU: 3D Medical Image Segmentation via Self-distilling TransUNet