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

Spinal surgery planning necessitates automatic segmentation of vertebrae incone-beam computed tomography (CBCT), an intraoperative imaging modality thatis widely used in intervention. However, CBCT images are of low-quality andartifact-laden due to noise, poor tissue contrast, and the presence of metallicobjects, causing vertebra segmentation, even manually, a demanding task. Incontrast, there exists a wealth of artifact-free, high quality CT images withvertebra annotations. This motivates us to build a CBCT vertebra segmentationmodel using unpaired CT images with annotations. To overcome the domain andartifact gaps between CBCT and CT, it is a must to address the threeheterogeneous tasks of vertebra segmentation, artifact reduction and modalitytranslation all together. To this, we propose a novel anatomy-aware artifactdisentanglement and segmentation network (A$^3$DSegNet) that intensivelyleverages knowledge sharing of these three tasks to promote learning.Specifically, it takes a random pair of CBCT and CT images as the input andmanipulates the synthesis and segmentation via different decoding combinationsfrom the disentangled latent layers. Then, by proposing various forms ofconsistency among the synthesized images and among segmented vertebrae, thelearning is achieved without paired (i.e., anatomically identical) data.Finally, we stack 2D slices together and build 3D networks on top to obtainfinal 3D segmentation result. Extensive experiments on a large number ofclinical CBCT (21,364) and CT (17,089) images show that the proposedA$^3$DSegNet performs significantly better than state-of-the-art competingmethods trained independently for each task and, remarkably, it achieves anaverage Dice coefficient of 0.926 for unpaired 3D CBCT vertebra segmentation.

A$^3$DSegNet: Anatomy-aware artifact disentanglement and segmentation  network for unpaired segmentation, artifact reduction, and modality  translation