Spatial segmentation of mass spectrometry imaging data featuring selected principal components.

Area of Science:

• Mass spectrometry imaging (MSI) data analysis
• Computational biology
• Bioinformatics

Background:

• Spatial segmentation in MSI is crucial for identifying spectral or ion image subgroups.
• Existing segmentation methods struggle with variations in data quality, resolution, and size.
• Dimensional reduction and clustering are fundamental but not universally optimal steps.

Purpose of the Study:

• To develop a robust and user-friendly spatial segmentation pipeline for MSI data.
• To improve the segmentation process by utilizing principal component analysis (PCA).
• To offer a readily integrable solution for biological analysis based on MSI data.

Main Methods:

• A novel segmentation pipeline utilizing pattern compression via principal component analysis (PCA).
• Representation of MSI data by normalized principal components for segmentation.
• Testing the pipeline on multiple rat brain and mouse kidney datasets.

Main Results:

• The proposed pipeline effectively segments MSI data using normalized principal components.
• Demonstrated advantages in ease of use and integration with existing pipelines.
• Successful application on diverse biological datasets (rat brains, mouse kidneys).

Conclusions:

• The PCA-based segmentation pipeline offers a significant advancement for MSI data analysis.
• Its ease of use and integration capabilities make it a valuable tool for researchers.
• Facilitates more effective biological analysis by providing reliable segmentation maps.