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 medical image segmentation, it is often necessary to collect opinions frommultiple experts to make the final decision. This clinical routine helps tomitigate individual bias. But when data is multiply annotated, standard deeplearning models are often not applicable. In this paper, we propose a novelneural network framework, called Multi-Rater Prism (MrPrism) to learn themedical image segmentation from multiple labels. Inspired by the iterativehalf-quadratic optimization, the proposed MrPrism will combine the multi-raterconfidences assignment task and calibrated segmentation task in a recurrentmanner. In this recurrent process, MrPrism can learn inter-observer variabilitytaking into account the image semantic properties, and finally converges to aself-calibrated segmentation result reflecting the inter-observer agreement.Specifically, we propose Converging Prism (ConP) and Diverging Prism (DivP) toprocess the two tasks iteratively. ConP learns calibrated segmentation based onthe multi-rater confidence maps estimated by DivP. DivP generates multi-raterconfidence maps based on the segmentation masks estimated by ConP. Theexperimental results show that by recurrently running ConP and DivP, the twotasks can achieve mutual improvement. The final converged segmentation resultof MrPrism outperforms state-of-the-art (SOTA) strategies on a wide range ofmedical image segmentation tasks.

Multi-rater Prism: Learning self-calibrated medical image segmentation from multiple raters