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

Imbalanced training data is a significant challenge for medical imageclassification. In this study, we propose a novel Progressive Class-CenterTriplet (PCCT) framework to alleviate the class imbalance issue particularlyfor diagnosis of rare diseases, mainly by carefully designing the tripletsampling strategy and the triplet loss formation. Specifically, the PCCTframework includes two successive stages. In the first stage, PCCT trains thediagnosis system via a class-balanced triplet loss to coarsely separatedistributions of different classes. In the second stage, the PCCT frameworkfurther improves the diagnosis system via a class-center involved triplet lossto cause a more compact distribution for each class. For the class-balancedtriplet loss, triplets are sampled equally for each class at each trainingiteration, thus alleviating the imbalanced data issue. For the class-centerinvolved triplet loss, the positive and negative samples in each triplet arereplaced by their corresponding class centers, which enforces datarepresentations of the same class closer to the class center. Furthermore, theclass-center involved triplet loss is extended to the pair-wise ranking lossand the quadruplet loss, which demonstrates the generalization of the proposedframework. Extensive experiments support that the PCCT framework workseffectively for medical image classification with imbalanced training images.On two skin image datasets and one chest X-ray dataset, the proposed approachrespectively obtains the mean F1 score 86.2, 65.2, and 90.66 over all classesand 81.4, 63.87, and 81.92 for rare classes, achieving state-of-the-artperformance and outperforming the widely used methods for the class imbalanceissue.

PCCT: Progressive Class-Center Triplet Loss for Imbalanced Medical Image Classification