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

Automatic segmentation methods are an important advancement in medical imageanalysis. Machine learning techniques, and deep neural networks in particular,are the state-of-the-art for most medical image segmentation tasks. Issues withclass imbalance pose a significant challenge in medical datasets, with lesionsoften occupying a considerably smaller volume relative to the background. Lossfunctions used in the training of deep learning algorithms differ in theirrobustness to class imbalance, with direct consequences for model convergence.The most commonly used loss functions for segmentation are based on either thecross entropy loss, Dice loss or a combination of the two. We propose theUnified Focal loss, a new hierarchical framework that generalises Dice andcross entropy-based losses for handling class imbalance. We evaluate ourproposed loss function on five publicly available, class imbalanced medicalimaging datasets: CVC-ClinicDB, Digital Retinal Images for Vessel Extraction(DRIVE), Breast Ultrasound 2017 (BUS2017), Brain Tumour Segmentation 2020(BraTS20) and Kidney Tumour Segmentation 2019 (KiTS19). We compare our lossfunction performance against six Dice or cross entropy-based loss functions,across 2D binary, 3D binary and 3D multiclass segmentation tasks, demonstratingthat our proposed loss function is robust to class imbalance and consistentlyoutperforms the other loss functions. Source code is available at:https://github.com/mlyg/unified-focal-loss

Unified Focal loss: Generalising Dice and cross entropy-based losses to handle class imbalanced medical image segmentation