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

High annotation costs and limited labels for dense 3D medical imaging taskshave recently motivated an assortment of 3D self-supervised pretraining methodsthat improve transfer learning performance. However, these methods commonlylack spatial awareness despite its centrality in enabling effective 3D imageanalysis. More specifically, position, scale, and orientation are not onlyinformative but also automatically available when generating image crops fortraining. Yet, to date, no work has proposed a pretext task that distills allkey spatial features. To fulfill this need, we develop a new self-supervisedmethod, VectorPOSE, which promotes better spatial understanding with two novelpretext tasks: Vector Prediction (VP) and Boundary-Focused Reconstruction(BFR). VP focuses on global spatial concepts (i.e., properties of 3D patches)while BFR addresses weaknesses of recent reconstruction methods to learn moreeffective local representations. We evaluate VectorPOSE on three 3D medicalimage segmentation tasks, showing that it often outperforms state-of-the-artmethods, especially in limited annotation settings.

A Point in the Right Direction: Vector Prediction for Spatially-aware Self-supervised Volumetric Representation Learning