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

3D reconstruction of medical imaging from 2D images has become anincreasingly interesting topic with the development of deep learning models inrecent years. Previous studies in 3D reconstruction from limited X-ray imagesmainly rely on learning from paired 2D and 3D images, where the reconstructionquality relies on the scale and variation of collected data. This has broughtsignificant challenges in the collection of training data, as only a tinyfraction of patients take two types of radiation examinations in the sameperiod. Although simulation from higher-dimension images could solve thisproblem, the variance between real and simulated data could bring greatuncertainty at the same time. In oral reconstruction, the situation becomesmore challenging as only a single panoramic X-ray image is available, wheremodels need to infer the curved shape by prior individual knowledge. Toovercome these limitations, we propose Oral-3Dv2 to solve this cross-dimensiontranslation problem in dental healthcare by learning solely on projectioninformation, i.e., the projection image and trajectory of the X-ray tube. Ourmodel learns to represent the 3D oral structure in an implicit way by mapping2D coordinates into density values of voxels in the 3D space. To improveefficiency and effectiveness, we utilize a multi-head model that predicts abunch of voxel values in 3D space simultaneously from a 2D coordinate in theaxial plane and the dynamic sampling strategy to refine details of the densitydistribution in the reconstruction result. Extensive experiments in simulatedand real data show that our model significantly outperforms existingstate-of-the-art models without learning from paired images or prior individualknowledge. To the best of our knowledge, this is the first work of anon-adversarial-learning-based model in 3D radiology reconstruction from asingle panoramic X-ray image.

Oral-3Dv2: 3D Oral Reconstruction from Panoramic X-Ray Imaging with Implicit Neural Representation