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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Diagnosing rare anemia disorders using microscopic images is challenging forskilled specialists and machine-learning methods alike. Due to thousands ofdisease-relevant cells in a single blood sample, this constitutes a complexmultiple-instance learning (MIL) problem. While the spatial neighborhood of redblood cells is not meaningful per se, the topology, i.e., the geometry of bloodsamples as a whole, contains informative features to remedy typical MIL issues,such as vanishing gradients and overfitting when training on limited data. Wethus develop a topology-based approach that extracts multi-scale topologicalfeatures from bags of single red blood cell images. The topological featuresare used to regularize the model, enforcing the preservation of characteristictopological properties of the data. Applied to a dataset of 71 patientssuffering from rare anemia disorders with 521 microscopic images of red bloodcells, our experiments show that topological regularization is an effectivemethod that leads to more than 3% performance improvements for the automatedclassification of rare anemia disorders based on single-cell images. This isthe first approach that uses topological properties for regularizing the MILprocess.

Topologically-Regularized Multiple Instance Learning for Red Blood Cell Disease Classification