Development of an object-based image analysis tool for mass spectrometry imaging ion classification

Area of Science:

• Analytical Chemistry
• Biotechnology
• Computational Biology

Background:

• Mass spectrometry imaging (MSI) generates complex datasets with thousands of molecular features.
• Distinguishing true biological signals from background noise in MSI data is a time-consuming manual process.
• Current manual methods for ion analysis are subjective and labor-intensive.

Purpose of the Study:

• To develop and validate an automated ion classification tool (ICT) for mass spectrometry imaging.
• To improve the efficiency and objectivity of distinguishing on-tissue analytes from background ions.
• To reduce the manual effort required for MSI data interpretation.

Main Methods:

• Development of an ion classification tool (ICT) using object-based image analysis in MATLAB.
• Segmentation of ion heatmap images into on-tissue and off-tissue regions via binary conversion.
• Classification of ions based on detected object counts using a binning approach.

Main Results:

• The ICT automates the classification of ions as on-tissue or background.
• The tool processes ion heatmaps within seconds, significantly reducing analysis time.
• In a test dataset, the ICT accurately classified 45 out of 50 randomly selected ions.

Conclusions:

• The developed ion classification tool (ICT) offers a rapid and objective method for MSI data analysis.
• This automation significantly reduces the time and subjectivity associated with manual ion discrimination.
• The ICT demonstrates high accuracy in classifying ions, enhancing the reliability of MSI studies.