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

The need for training data can impede the adoption of novel imagingmodalities for learning-based medical image analysis. Domain adaptation methodspartially mitigate this problem by translating training data from a relatedsource domain to a novel target domain, but typically assume that a one-to-onetranslation is possible. Our work addresses the challenge of adapting to a moreinformative target domain where multiple target samples can emerge from asingle source sample. In particular we consider translating from mp-MRI toVERDICT, a richer MRI modality involving an optimized acquisition protocol forcancer characterization. We explicitly account for the inherent uncertainty ofthis mapping and exploit it to generate multiple outputs conditioned on asingle input. Our results show that this allows us to extract systematicallybetter image representations for the target domain, when used in tandem withboth simple, CycleGAN-based baselines, as well as more powerful approaches thatintegrate discriminative segmentation losses and/or residual adapters. Whencompared to its deterministic counterparts, our approach yields substantialimprovements across a broad range of dataset sizes, increasingly strongbaselines, and evaluation measures.

Harnessing Uncertainty in Domain Adaptation for MRI Prostate Lesion Segmentation