ShapeMetrics: A userfriendly pipeline for 3D cell segmentation and spatial tissue analysis

Area of Science:

• Life Sciences
• Computational Biology
• Bioimaging

Background:

• Increasing demand for single-cell data in life sciences.
• Current 3D cell segmentation methods struggle with complex in vivo tissues.
• Need for robust tools for volumetric and morphological analysis of cells in 3D.

Purpose of the Study:

• To develop a user-friendly MATLAB-based script, ShapeMetrics, for 3D cell segmentation.
• To enable unbiased clustering and classification of cells based on volumetric and morphological features.
• To facilitate the integration of machine learning-based analysis with traditional biomarker data.

Main Methods:

• Generation of ShapeMetrics, a MATLAB script for 3D cell segmentation.
• Application of unbiased clustering using heatmaps for cell subgrouping.
• Analysis of volumetric and morphological features including ellipticity, longest axis, and volume-to-surface area ratio.
• Spatial mapping of segmented cells back to their original tissue locations.

Main Results:

• ShapeMetrics accurately segments cells in 3D images.
• The script effectively segregates cells into subgroups based on diverse morphological and volumetric characteristics.
• Machine learning integration allows for novel data dissection beyond fluorescent biomarkers.
• Spatial mapping provides context for cell function within tissue morphology.

Conclusions:

• ShapeMetrics offers a user-friendly solution for single-cell level accuracy in 3D imaging analysis.
• The method enhances the dissection of complex biological data from microscope images.
• It provides valuable insights into cell morphology and spatial organization within tissues.
• Facilitates transition from bulk to single-cell analysis for researchers with limited computational biology experience.