CSMM

CSMM computes cosine similarity measure maps from mass spectrometry imaging (MSI) pixels to quantify spatial molecular similarity and map tissue heterogeneity.

Key Features:

• Cosine similarity mapping: Calculates cosine similarity between a selected MSI pixel and all other image pixels to generate CSMM maps.
• Spatial visualization of similarity metrics: Represents spatial distributions of cosine similarity values to reveal tissue heterogeneity and area borders within samples.
• Sensitivity to reference pixel selection: Enables examination of how altering the reference CSMM pixel affects similarity distributions and boundary detection.
• Quantitative heterogeneity assessment: Provides quantitative measures of molecular similarity across the image without relying on predefined clusters or borders.

Scientific Applications:

• Tissue heterogeneity analysis: Mapping intra-sample molecular heterogeneity to identify distinct tissue types and distribution patterns in MSI data.
• Boundary delineation in complex samples: Detecting area borders and spatial transitions based on molecular similarity distributions.
• Complement to histology and diagnostics: Providing spatial molecular information to complement traditional histology and support diagnostic studies.

Methodology:

Computes cosine similarity metrics between a chosen MSI reference pixel and other image pixels to produce CSMM maps, yielding quantitative spatial similarity measures used to identify heterogeneity patterns and borders without predefined clustering.