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

Brain tumor segmentation (BTS) in magnetic resonance image (MRI) is crucialfor brain tumor diagnosis, cancer management and research purposes. With thegreat success of the ten-year BraTS challenges as well as the advances of CNNand Transformer algorithms, a lot of outstanding BTS models have been proposedto tackle the difficulties of BTS in different technical aspects. However,existing studies hardly consider how to fuse the multi-modality images in areasonable manner. In this paper, we leverage the clinical knowledge of howradiologists diagnose brain tumors from multiple MRI modalities and propose aclinical knowledge-driven brain tumor segmentation model, called CKD-TransBTS.Instead of directly concatenating all the modalities, we re-organize the inputmodalities by separating them into two groups according to the imagingprinciple of MRI. A dual-branch hybrid encoder with the proposedmodality-correlated cross-attention block (MCCA) is designed to extract themulti-modality image features. The proposed model inherits the strengths fromboth Transformer and CNN with the local feature representation ability forprecise lesion boundaries and long-range feature extraction for 3D volumetricimages. To bridge the gap between Transformer and CNN features, we propose aTrans&CNN Feature Calibration block (TCFC) in the decoder. We compare theproposed model with five CNN-based models and six transformer-based models onthe BraTS 2021 challenge dataset. Extensive experiments demonstrate that theproposed model achieves state-of-the-art brain tumor segmentation performancecompared with all the competitors.

CKD-TransBTS: Clinical Knowledge-Driven Hybrid Transformer with Modality-Correlated Cross-Attention for Brain Tumor Segmentation