MIRIAM

MIRIAM performs single-cell segmentation and quantification in highly multiplexed tissue imaging to enable accurate protein expression profiling and morphological analysis.

Key Features:

• Machine learning-based pixel classification: Performs pixel-level classification to define and segment cellular compartments within tissue images.
• Method for extending incomplete cell membranes: Extends interrupted or incomplete cell membranes to improve contiguous cell boundaries for segmentation.
• Deep learning-based cell shape descriptor: Generates learned descriptors of cell morphology using deep learning to capture intricate cellular shapes.

Scientific Applications:

• Multiplexed tissue imaging single-cell analysis: Enables precise single-cell protein expression profiling and morphological analysis across highly multiplexed tissue imaging platforms.
• Tumor tissue analysis (human colonic adenomas): Applied to segmentation and quantification of cells in human colonic adenomas for tissue biology studies.

Methodology:

The pipeline integrates machine learning-based pixel classification, a method for extending incomplete cell membranes, and a deep learning-based cell shape descriptor for segmentation and quantification.