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

Strong speckle noise is inherent to optical coherence tomography (OCT)imaging and represents a significant obstacle for accurate quantitativeanalysis of retinal structures which is key for advances in clinical diagnosisand monitoring of disease. Learning-based self-supervised methods forstructure-preserving noise reduction have demonstrated superior performanceover traditional methods but face unique challenges in OCT imaging. The highcorrelation of voxels generated by coherent A-scan beams undermines theefficacy of self-supervised learning methods as it violates the assumption ofindependent pixel noise. We conduct experiments demonstrating limitations ofexisting models due to this independence assumption. We then introduce a newend-to-end self-supervised learning framework specifically tailored for OCTimage denoising, integrating slice-by-slice training and registration modulesinto one network. An extensive ablation study is conducted for the proposedapproach. Comparison to previously published self-supervised denoising modelsdemonstrates improved performance of the proposed framework, potentiallyserving as a preprocessing step towards superior segmentation performance andquantitative analysis.

Self-supervised OCT Image Denoising with Slice-to-Slice Registration and Reconstruction