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

Generalizability is an important problem in deep neural networks, especiallyin the context of the variability of data acquisition in clinical magneticresonance imaging (MRI). Recently, the Spatially Localized Atlas Network Tiles(SLANT) approach has been shown to effectively segment whole brain non-contrastT1w MRI with 132 volumetric labels. Enhancing generalizability of SLANT wouldenable broader application of volumetric assessment in multi-site studies.Transfer learning (TL) is commonly used to update the neural network weightsfor local factors; yet, it is commonly recognized to risk degradation ofperformance on the original validation/test cohorts. Here, we explore TL bydata augmentation to address these concerns in the context of adapting SLANT toanatomical variation and scanning protocol. We consider two datasets: First, weoptimize for age with 30 T1w MRI of young children with manually correctedvolumetric labels, and accuracy of automated segmentation defined relative tothe manually provided truth. Second, we optimize for acquisition with 36 paireddatasets of pre- and post-contrast clinically acquired T1w MRI, and accuracy ofthe post-contrast segmentations assessed relative to the pre-contrast automatedassessment. For both studies, we augment the original TL step of SLANT witheither only the new data or with both original and new data. Over baselineSLANT, both approaches yielded significantly improved performance (signed ranktests; pediatric: 0.89 vs. 0.82 DSC, p<0.001; contrast: 0.80 vs 0.76, p<0.001).The performance on the original test set decreased with the new-data onlytransfer learning approach, so data augmentation was superior to stricttransfer learning.

Generalizing Deep Whole Brain Segmentation for Pediatric and Post-Contrast MRI with Augmented Transfer Learning