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 Open

Presents the pdf analysis as premium feature

PDF Analysis

Insights

Presents the summarisation as premium feature

Summarisation

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the zaia usage as premium feature

ZAIA

Zendy Tools

Zendy Plus

Presents the access of premium content as premium feature

Premium Content

State-of-the-art brain tumor segmentation is based on deep learning modelsapplied to multi-modal MRIs. Currently, these models are trained on imagesafter a preprocessing stage that involves registration, interpolation, brainextraction (BE, also known as skull-stripping) and manual correction by anexpert. However, for clinical practice, this last step is tedious andtime-consuming and, therefore, not always feasible, resulting inskull-stripping faults that can negatively impact the tumor segmentationquality. Still, the extent of this impact has never been measured for any ofthe many different BE methods available. In this work, we propose an automaticbrain tumor segmentation pipeline and evaluate its performance with multiple BEmethods. Our experiments show that the choice of a BE method can compromise upto 15.7% of the tumor segmentation performance. Moreover, we propose trainingand testing tumor segmentation models on non-skull-stripped images, effectivelydiscarding the BE step from the pipeline. Our results show that this approachleads to a competitive performance at a fraction of the time. We conclude that,in contrast to the current paradigm, training tumor segmentation models onnon-skull-stripped images can be the best option when high performance inclinical practice is desired.

Towards fully automated deep-learning-based brain tumor segmentation: is brain extraction still necessary?