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

Deep learning models with large learning capacities often overfit to medicalimaging datasets. This is because training sets are often relatively small dueto the significant time and financial costs incurred in medical dataacquisition and labelling. Data augmentation is therefore often used to expandthe availability of training data and to increase generalization. However,augmentation strategies are often chosen on an ad-hoc basis withoutjustification. In this paper, we present an augmentation policy search methodwith the goal of improving model classification performance. We include in theaugmentation policy search additional transformations that are often used inmedical image analysis and evaluate their performance. In addition, we extendthe augmentation policy search to include non-linear mixed-example dataaugmentation strategies. Using these learned policies, we show that principleddata augmentation for medical image model training can lead to significantimprovements in ultrasound standard plane detection, with an an averageF1-score improvement of 7.0% overall over naive data augmentation strategies inultrasound fetal standard plane classification. We find that the learnedrepresentations of ultrasound images are better clustered and defined withoptimized data augmentation.

Principled Ultrasound Data Augmentation for Classification of Standard Planes