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

White matter (WM) tract segmentation is a crucial step for brain connectivitystudies. It is performed on diffusion magnetic resonance imaging (dMRI), anddeep neural networks (DNNs) have achieved promising segmentation accuracy.Existing DNN-based methods use an annotated dataset for model training.However, the performance of the trained model on a different test dataset maynot be optimal due to distribution shift, and it is desirable to design WMtract segmentation approaches that allow better generalization of thesegmentation model to arbitrary test datasets. In this work, we propose a WMtract segmentation approach that improves the generalization with scaledresidual bootstrap. The difference between dMRI scans in training and testdatasets is most noticeably caused by the different numbers of diffusiongradients and noise levels. Since both of them lead to differentsignal-to-noise ratios (SNRs) between the training and test data, we propose toaugment the training scans by adjusting the noise magnitude and develop anadapted residual bootstrap strategy for the augmentation. To validate theproposed approach, two dMRI datasets were used, and the experimental resultsshow that our method consistently improved the generalization of WM tractsegmentation under various settings.

Better Generalization of White Matter Tract Segmentation to Arbitrary Datasets with Scaled Residual Bootstrap