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

Assessing the degree of disease severity in biomedical images is a tasksimilar to standard classification but constrained by an underlying structurein the label space. Such a structure reflects the monotonic relationshipbetween different disease grades. In this paper, we propose a straightforwardapproach to enforce this constraint for the task of predicting DiabeticRetinopathy (DR) severity from eye fundus images based on the well-known notionof Cost-Sensitive classification. We expand standard classification losses withan extra term that acts as a regularizer, imposing greater penalties onpredicted grades when they are farther away from the true grade associated to aparticular image. Furthermore, we show how to adapt our method to the modellingof label noise in each of the sub-problems associated to DR grading, anapproach we refer to as Atomic Sub-Task modeling. This yields models that canimplicitly take into account the inherent noise present in DR gradeannotations. Our experimental analysis on several public datasets reveals that,when a standard Convolutional Neural Network is trained using this simplestrategy, improvements of 3-5\% of quadratic-weighted kappa scores can beachieved at a negligible computational cost. Code to reproduce our results isreleased at https://github.com/agaldran/cost_sensitive_loss_classification.

Cost-Sensitive Regularization for Diabetic Retinopathy Grading from Eye Fundus Images