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/

Zendy Plus

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

Zendy Tools

Zendy Open

Deep learning approaches to breast cancer detection in mammograms haverecently shown promising results. However, such models are constrained by thelimited size of publicly available mammography datasets, in large part due toprivacy concerns and the high cost of generating expert annotations. Limiteddataset size is further exacerbated by substantial class imbalance since"normal" images dramatically outnumber those with findings. Given the rapidprogress of generative models in synthesizing realistic images, and the knowneffectiveness of simple data augmentation techniques (e.g. horizontalflipping), we ask if it is possible to synthetically augment mammogram datasetsusing generative adversarial networks (GANs). We train a class-conditional GANto perform contextual in-filling, which we then use to synthesize lesions ontohealthy screening mammograms. First, we show that GANs are capable ofgenerating high-resolution synthetic mammogram patches. Next, we experimentallyevaluate using the augmented dataset to improve breast cancer classificationperformance. We observe that a ResNet-50 classifier trained with GAN-augmentedtraining data produces a higher AUROC compared to the same model trained onlyon traditionally augmented data, demonstrating the potential of our approach.

Conditional Infilling GANs for Data Augmentation in Mammogram Classification