Image-based quantification of histological features as a function of spatial location using the Tissue Positioning...

Area of Science:

• Spatial biology
• Genomics
• Computational pathology

Background:

• Tissues exhibit complex gene expression patterns crucial for cell function.
• Quantifying these zonal patterns requires precise measurement of cellular features along spatial axes.
• Manual analysis is labor-intensive and prone to inaccuracies.

Purpose of the Study:

• To develop an automated method for analyzing tissue zonation.
• To enable precise spatial quantification of gene expression in relation to zonal architecture.
• To overcome limitations of manual analysis in studying zonal biology.

Main Methods:

• Development of the deep-learning-based Tissue Positioning System (TPS).
• TPS utilizes algorithms for vessel identification, classification, and zone segmentation.
• Analysis is performed along the portal vein to central vein axis in the liver lobule.

Main Results:

• TPS accurately quantifies spatial gene expression in mouse models.
• The system successfully identified expression differences across zonal reporter strains, ages, and disease states.
• TPS revealed the zonal distribution of previously assumed randomly located cells.

Conclusions:

• The Tissue Positioning System (TPS) offers an automated and accurate approach to studying tissue zonation.
• TPS can differentiate biological variations and uncover novel spatial cell distributions.
• The principles of TPS can be extended to other tissues for broader applications in spatial biology.