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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Interactive segmentation reduces the annotation time of medical images andallows annotators to iteratively refine labels with corrective interactions,such as clicks. While existing interactive models transform clicks into userguidance signals, which are combined with images to form (image, guidance)pairs, the question of how to best represent the guidance has not been fullyexplored. To address this, we conduct a comparative study of existing guidancesignals by training interactive models with different signals and parametersettings to identify crucial parameters for the model's design. Based on ourfindings, we design a guidance signal that retains the benefits of othersignals while addressing their limitations. We propose an adaptive Gaussianheatmaps guidance signal that utilizes the geodesic distance transform todynamically adapt the radius of each heatmap when encoding clicks. We conductour study on the MSD Spleen and the AutoPET datasets to explore thesegmentation of both anatomy (spleen) and pathology (tumor lesions). Ourresults show that choosing the guidance signal is crucial for interactivesegmentation as we improve the performance by 14% Dice with our adaptiveheatmaps on the challenging AutoPET dataset when compared to non-interactivemodels. This brings interactive models one step closer to deployment onclinical workflows. We will make our code publically available.

Guiding the Guidance: A Comparative Analysis of User Guidance Signals for Interactive Segmentation of Volumetric Images