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

Task-specific deep learning models in histopathology offer promisingopportunities for improving diagnosis, clinical research, and precisionmedicine. However, development of such models is often limited by availabilityof high-quality data. Foundation models in histopathology that learn generalrepresentations across a wide range of tissue types, diagnoses, andmagnifications offer the potential to reduce the data, compute, and technicalexpertise necessary to develop task-specific deep learning models with therequired level of model performance. In this work, we describe the developmentand evaluation of foundation models for histopathology via self-supervisedlearning (SSL). We first establish a diverse set of benchmark tasks involving17 unique tissue types and 12 unique cancer types and spanning differentoptimal magnifications and task types. Next, we use this benchmark to exploreand evaluate histopathology-specific SSL methods followed by further evaluationon held out patch-level and weakly supervised tasks. We found that standard SSLmethods thoughtfully applied to histopathology images are performant across ourbenchmark tasks and that domain-specific methodological improvements canfurther increase performance. Our findings reinforce the value of usingdomain-specific SSL methods in pathology, and establish a set of high qualityfoundation models to enable further research across diverse applications.

Domain-specific optimization and diverse evaluation of self-supervised models for histopathology