Semi-automated approaches for interrogating spatial heterogeneity of tissue samples

Area of Science:

• Spatial biology
• Biomedical imaging
• Computational pathology

Background:

• Tissues are complex ecosystems where cellular interactions drive biological processes and disease.
• Understanding tissue composition at single-cell resolution is crucial for biological and medical insights.
• Multiplex whole-tissue imaging generates large datasets requiring advanced analytical methods.

Purpose of the Study:

• To present a semi-automated workflow for analyzing multiplex whole-tissue images.
• To demonstrate the utility of open-source tools for spatial biology data extraction.
• To facilitate the exploration of tumor microenvironment (TME) composition at single-cell resolution.

Main Methods:

• Utilized the Lunaphore COMET platform for interrogating lung cancer specimens with 20 biomarkers.
• Employed an in-house optimized nuclei detection algorithm and a novel image artifact exclusion approach.
• Processed data using publicly available image analysis frameworks to ensure compatibility.

Main Results:

• Demonstrated a semi-automated workflow for multiplex whole-tissue image analysis.
• Showcased the compatibility of COMET-derived data with existing analysis frameworks.
• Enabled single-cell resolution spatial cellular dissection of tissue composition.

Conclusions:

• The presented workflow offers an accessible, slide-in, data-out approach for multiplex imaging analysis.
• This method simplifies the exploration of TME composition at single-cell resolution.
• The workflow is transferable to diverse specimen cohorts, providing a valuable toolset for spatial biology research.