Related Experiment Video
Updated: May 5, 2026
08:59
Morphology-Based Distinction Between Healthy and Pathological Cells Utilizing Fourier Transforms and Self-Organizing Maps
Published on: October 28, 2018
8.7K
Self-organizing maps: a versatile tool for the automatic analysis of untargeted imaging datasets
Pietro Franceschi1, Ron Wehrens
1Biostatistics and Data Management, IASMA Research and Innovation Centre, Fondazione E. Mach, Trento, Italy.
Proteomics
|November 26, 2013
Summary
This study introduces a new method using self-organizing maps for analyzing large mass spectrometry imaging datasets. The approach generates spatial ion distribution images, simplifying the analysis of complex biological samples.
Area of Science:
- Biochemistry
- Analytical Chemistry
- Data Science
Background:
- Mass spectrometry (MS)-based imaging enables location-specific chemical identification in biological samples, enhancing understanding of biological processes.
- Analyzing the massive datasets generated by high-resolution MS imaging presents significant computational challenges.
- Existing methods struggle with the high dimensionality and volume of data from modern MS imaging techniques.
Purpose of the Study:
- To develop a novel computational approach for analyzing large-scale mass spectrometry imaging data.
- To address the challenges associated with data analysis in high-resolution mass spectrometry imaging.
- To enable more detailed insights into biological mechanisms through efficient data processing.
Main Methods:
- A novel approach based on self-organizing maps (SOMs) is presented, extending previous work to handle high-resolution mass spectra.
- The core innovation involves generating prototype images of ion spatial distributions instead of prototypical mass spectra.
- A two-stage analysis is employed: first, identifying typical spatial distributions and associated m/z bins, followed by detailed analysis of selected bins using accurate masses.
Main Results:
- The proposed method successfully handles the large number of variables inherent in high-resolution mass spectra.
- Prototype images representing spatial ion distributions were generated, facilitating a new analytical pathway.
- The approach was validated using an in-house dataset of apple slices, demonstrating its practical applicability.
Conclusions:
- The novel self-organizing map-based approach offers an effective solution for analyzing large and complex mass spectrometry imaging datasets.
- Generating prototype images of ion distributions significantly simplifies the analysis of high-resolution mass spectra.
- This method enhances the potential of MS imaging for detailed biological and chemical investigations.
