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

Self-supervised learning (SSL) enables label efficient training for machinelearning models. This is essential for domains such as medical imaging, wherelabels are costly and time-consuming to curate. However, the most effectivesupervised or SSL strategy for transferring models to different healthcaresystems or novel tasks is not well understood. In this work, we systematicallyexperiment with a variety of supervised and self-supervised pretrainingstrategies using multimodal datasets of medical images (chest X-rays) and text(radiology reports). We then evaluate their performance on data from twoexternal institutions with diverse sets of tasks. In addition, we experimentwith different transfer learning strategies to effectively adapt thesepretrained models to new tasks and healthcare systems. Our empirical resultssuggest that multimodal SSL gives substantial gains over unimodal SSL inperformance across new healthcare systems and tasks, comparable to modelspretrained with full supervision. We demonstrate additional performance gainswith models further adapted to the new dataset and task, using multimodaldomain-adaptive pretraining (DAPT), linear probing then finetuning (LP-FT), andboth methods combined. We offer suggestions for alternative models to use inscenarios where not all of these additions are feasible. Our results provideguidance for improving the generalization of medical image interpretationmodels to new healthcare systems and novel tasks.

How to Train Your CheXDragon: Training Chest X-Ray Models for Transfer to Novel Tasks and Healthcare Systems