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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Is self-supervised deep learning (DL) for medical image analysis already aserious alternative to the de facto standard of end-to-end trained supervisedDL? We tackle this question for medical image classification, with a particularfocus on one of the currently most limiting factors of the field: the(non-)availability of labeled data. Based on three common medical imagingmodalities (bone marrow microscopy, gastrointestinal endoscopy, dermoscopy) andpublicly available data sets, we analyze the performance of self-supervised DLwithin the self-distillation with no labels (DINO) framework. After learning animage representation without use of image labels, conventional machine learningclassifiers are applied. The classifiers are fit using a systematically variednumber of labeled data (1-1000 samples per class). Exploiting the learned imagerepresentation, we achieve state-of-the-art classification performance for allthree imaging modalities and data sets with only a fraction of between 1% and10% of the available labeled data and about 100 labeled samples per class.

Self-supervision for medical image classification: state-of-the-art performance with ~100 labeled training samples per class