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

Dense prediction tasks such as segmentation and detection of pathologicalentities hold crucial clinical value in computational pathology workflows.However, obtaining dense annotations on large cohorts is usually tedious andexpensive. Contrastive learning (CL) is thus often employed to leverage largevolumes of unlabeled data to pre-train the backbone network. To boost CL fordense prediction, some studies have proposed variations of dense matchingobjectives in pre-training. However, our analysis shows that employing existingdense matching strategies on histopathology images enforces invariance amongincorrect pairs of dense features and, thus, is imprecise. To address this, wepropose a precise location-based matching mechanism that utilizes theoverlapping information between geometric transformations to precisely matchregions in two augmentations. Extensive experiments on two pretraining datasets(TCGA-BRCA, NCT-CRC-HE) and three downstream datasets (GlaS, CRAG, BCSS)highlight the superiority of our method in semantic and instance segmentationtasks. Our method outperforms previous dense matching methods by up to 7.2% inaverage precision for detection and 5.6% in average precision for instancesegmentation tasks. Additionally, by using our matching mechanism in the threepopular contrastive learning frameworks, MoCo-v2, VICRegL, and ConCL, theaverage precision in detection is improved by 0.7% to 5.2%, and the averageprecision in segmentation is improved by 0.7% to 4.0%, demonstratinggeneralizability. Our code is available athttps://github.com/cvlab-stonybrook/PLM_SSL.

Precise Location Matching Improves Dense Contrastive Learning in Digital Pathology