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

Various deep learning models have been developed to segment anatomicalstructures from medical images, but they typically have poor performance whentested on another target domain with different data distribution. Recently,unsupervised domain adaptation methods have been proposed to alleviate thisso-called domain shift issue, but most of them are designed for scenarios withrelatively small domain shifts and are likely to fail when encountering a largedomain gap. In this paper, we propose DCDA, a novel cross-modality unsuperviseddomain adaptation framework for tasks with large domain shifts, e.g.,segmenting retinal vessels from OCTA and OCT images. DCDA mainly consists of adisentangling representation style transfer (DRST) module and a collaborativeconsistency learning (CCL) module. DRST decomposes images into contentcomponents and style codes and performs style transfer and imagereconstruction. CCL contains two segmentation models, one for source domain andthe other for target domain. The two models use labeled data (together with thecorresponding transferred images) for supervised learning and performcollaborative consistency learning on unlabeled data. Each model focuses on thecorresponding single domain and aims to yield an expertized domain-specificsegmentation model. Through extensive experiments on retinal vesselsegmentation, our framework achieves Dice scores close to target-trained oracleboth from OCTA to OCT and from OCT to OCTA, significantly outperforming otherstate-of-the-art methods.

Unsupervised Domain Adaptation for Cross-Modality Retinal Vessel  Segmentation via Disentangling Representation Style Transfer and  Collaborative Consistency Learning