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

The recent state-of-the-art deep learning methods have significantly improvedbrain tumor segmentation. However, fully supervised training requires a largeamount of manually labeled masks, which is highly time-consuming and needsdomain expertise. Weakly supervised learning with scribbles provides a goodtrade-off between model accuracy and the effort of manual labeling. However,for segmenting the hierarchical brain tumor structures, manually labelingscribbles for each substructure could still be demanding. In this paper, we useonly two kinds of weak labels, i.e., scribbles on whole tumor and healthy braintissue, and global labels for the presence of each substructure, to train adeep learning model to segment all the sub-regions. Specifically, we train twonetworks in two phases: first, we only use whole tumor scribbles to train awhole tumor (WT) segmentation network, which roughly recovers the WT mask oftraining data; then we cluster the WT region with the guide of global labels.The rough substructure segmentation from clustering is used as weak labels totrain the second network. The dense CRF loss is used to refine the weaklysupervised segmentation. We evaluate our approach on the BraTS2017 dataset andachieve competitive WT dice score as well as comparable scores on substructuresegmentation compared to an upper bound when trained with fully annotatedmasks.

Scribble-based Hierarchical Weakly Supervised Learning for Brain Tumor Segmentation