Fig. 1: SpatialTopic unsupervisedly identifies distinct tissue microenvironments across images, utilizing topic model concepts in computer vision. | Nature Communications

Fig. 1: SpatialTopic unsupervisedly identifies distinct tissue microenvironments across images, utilizing topic model concepts in computer vision.

From: Scalable topic modelling decodes spatial tissue architecture for large-scale multiplexed imaging analysis

Fig. 1: SpatialTopic unsupervisedly identifies distinct tissue microenvironments across images, utilizing topic model concepts in computer vision.The alternative text for this image may have been generated using AI.

A Overview of SpatialTopic. SpatialTopic identifies biologically relevant topics across multiple images, while each topic is a distribution of cell types, reflecting the spatial tissue architecture across images. B Image analysis pipeline designed for multiplexed immunofluorescence images. SpatialTopic is designed as a critical step for spatial analysis after cell phenotyping. C Graphic representation for SpatialTopic. D SpatialTopic groups cells in an unsupervised manner based on spatially co-occurrent cell types, similar to image segmentation based on spatially co-occurrent colors in the photo.

Back to article page