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

In biomedical image analysis, the applicability of deep learning methods isdirectly impacted by the quantity of image data available. This is due to deeplearning models requiring large image datasets to provide high-levelperformance. Generative Adversarial Networks (GANs) have been widely utilizedto address data limitations through the generation of synthetic biomedicalimages. GANs consist of two models. The generator, a model that learns how toproduce synthetic images based on the feedback it receives. The discriminator,a model that classifies an image as synthetic or real and provides feedback tothe generator. Throughout the training process, a GAN can experience severaltechnical challenges that impede the generation of suitable synthetic imagery.First, the mode collapse problem whereby the generator either produces anidentical image or produces a uniform image from distinct input features.Second, the non-convergence problem whereby the gradient descent optimizerfails to reach a Nash equilibrium. Thirdly, the vanishing gradient problemwhereby unstable training behavior occurs due to the discriminator achievingoptimal classification performance resulting in no meaningful feedback beingprovided to the generator. These problems result in the production of syntheticimagery that is blurry, unrealistic, and less diverse. To date, there has beenno survey article outlining the impact of these technical challenges in thecontext of the biomedical imagery domain. This work presents a review andtaxonomy based on solutions to the training problems of GANs in the biomedicalimaging domain. This survey highlights important challenges and outlines futureresearch directions about the training of GANs in the domain of biomedicalimagery.

A Survey on Training Challenges in Generative Adversarial Networks for Biomedical Image Analysis