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

Domain generalization in medical image classification is an important problemfor trustworthy machine learning to be deployed in healthcare. We find thatexisting approaches for domain generalization which utilize ground-truthabnormality segmentations to control feature attributions have poorout-of-distribution (OOD) performance relative to the standard baseline ofempirical risk minimization (ERM). We investigate what regions of an image areimportant for medical image classification and show that parts of thebackground, that which is not contained in the abnormality segmentation,provides helpful signal. We then develop a new task-specific mask which coversall relevant regions. Utilizing this new segmentation mask significantlyimproves the performance of the existing methods on the OOD test sets. Toobtain better generalization results than ERM, we find it necessary to scale upthe training data size in addition to the usage of these task-specific masks.

The Importance of Background Information for Out of Distribution Generalization