Correlative Hyperspectral Imaging Using a Dimensionality-Reduction-Based Image Fusion Method.

Area of Science:

• Analytical Chemistry
• Spectroscopy
• Materials Science

Background:

• Chemical imaging techniques, particularly mass spectrometry imaging (MSI), are increasingly combined to enhance sample analysis.
• Multiple ionization sources in MSI enable detection of diverse molecular classes at various spatial resolutions.
• Current data fusion methods often rely on image registration, which can introduce interpolation artifacts.

Purpose of the Study:

• To present a novel data fusion method for chemical imaging that circumvents interpolation-based artifacts.
• To apply this method to data from multiple chemical imaging modalities.
• To evaluate the method's performance using simulated and real-world samples.

Main Methods:

• Development of a new data fusion algorithm for chemical imaging data.
• Application of the method to data acquired from multiple chemical imaging techniques.
• Validation using simulated data, a polymer blend model, and biological samples (mouse brain and lung).

Main Results:

• The proposed method successfully fuses data from multiple chemical imaging modalities without introducing interpolation artifacts.
• Evaluation with simulated and model samples demonstrated the effectiveness of the approach.
• Successful application to complex biological samples like mouse brain and lung tissues.

Conclusions:

• The developed data fusion method offers an artifact-free approach to combining chemical imaging data.
• This technique provides enhanced molecular information compared to individual methods or traditional fusion approaches.
• The method holds significant potential for advancing the analysis of complex biological and material samples.