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

Diabetic Retinopathy (DR) is a leading cause of vision loss in the world, andearly DR detection is necessary to prevent vision loss and support anappropriate treatment. In this work, we leverage interactive machine learningand introduce a joint learning framework, termed DRG-Net, to effectively learnboth disease grading and multi-lesion segmentation. Our DRG-Net consists of twomodules: (i) DRG-AI-System to classify DR Grading, localize lesion areas, andprovide visual explanations; (ii) DRG-Expert-Interaction to receive feedbackfrom user-expert and improve the DRG-AI-System. To deal with sparse data, weutilize transfer learning mechanisms to extract invariant featurerepresentations by using Wasserstein distance and adversarial learning-basedentropy minimization. Besides, we propose a novel attention strategy at bothlow- and high-level features to automatically select the most significantlesion information and provide explainable properties. In terms of humaninteraction, we further develop DRG-Net as a tool that enables expert users tocorrect the system's predictions, which may then be used to update the systemas a whole. Moreover, thanks to the attention mechanism and loss functionsconstraint between lesion features and classification features, our approachcan be robust given a certain level of noise in the feedback of users. We havebenchmarked DRG-Net on the two largest DR datasets, i.e., IDRID and FGADR, andcompared it to various state-of-the-art deep learning networks. In addition tooutperforming other SOTA approaches, DRG-Net is effectively updated using userfeedback, even in a weakly-supervised manner.

DRG-Net: Interactive Joint Learning of Multi-lesion Segmentation and Classification for Diabetic Retinopathy Grading