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

Though, deep learning based medical image registration is currently startingto show promising advances, often, it still fells behind conventionalframeworks in terms of registration accuracy. This is especially true forapplications where large deformations exist, such as registration ofinterpatient abdominal MRI or inhale-to-exhale CT lung registration. Mostcurrent works use U-Net-like architectures to predict dense displacement fieldsfrom the input images in different supervised and unsupervised settings. Webelieve that the U-Net architecture itself to some level limits the ability topredict large deformations (even when using multilevel strategies) andtherefore propose a novel approach, where the input images are mapped into adisplacement space and final registrations are reconstructed from thisembedding. Experiments on inhale-to-exhale CT lung registration demonstrate theability of our architecture to predict large deformations in a single forwardpath through our network (leading to errors below 2 mm).

Tackling the Problem of Large Deformations in Deep Learning Based Medical Image Registration Using Displacement Embeddings