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

Artificial intelligence(AI)-assisted method had received much attention inthe risk field such as disease diagnosis. Different from the classification ofdisease types, it is a fine-grained task to classify the medical images asbenign or malignant. However, most research only focuses on improving thediagnostic accuracy and ignores the evaluation of model reliability, whichlimits its clinical application. For clinical practice, calibration presentsmajor challenges in the low-data regime extremely for over-parametrized modelsand inherent noises. In particular, we discovered that modeling data-dependentuncertainty is more conducive to confidence calibrations. Compared withtest-time augmentation(TTA), we proposed a modified Bootstrapping loss(BS loss)function with Mixup data augmentation strategy that can better calibratepredictive uncertainty and capture data distribution transformation withoutadditional inference time. Our experiments indicated that BS loss withMixup(BSM) model can halve the Expected Calibration Error(ECE) compared tostandard data augmentation, deep ensemble and MC dropout. The correlationbetween uncertainty and similarity of in-domain data is up to -0.4428 under theBSM model. Additionally, the BSM model is able to perceive the semanticdistance of out-of-domain data, demonstrating high potential in real-worldclinical practice.

BSM loss: A superior way in modeling aleatory uncertainty of fine_grained classification