PatternExtract: A facile, scalable pipeline for point pattern generation from spatial imaging data

Read the full article See related articles

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

We present PatternExtract, an open-source pipeline that generates accurate spatial point patterns from RGB pathology images and cell coordinate data without relying on composite channels or proprietary software. Using a novel two-kernel tissue segmentation method combined with automated pixel classification in QuPath, PatternExtract precisely excludes tissue artifacts such as necrosis and blood vessels to define spatial windows. Optimized on 568 diffuse large B-cell lymphoma images and validated on an independent cohort, the pipeline enables robust spatial analyses.

Article activity feed