Assessment of Essential Information in the Fourier Domain to Accelerate Raman Hyperspectral Microimaging

Area of Science:

• Chemometrics
• Spectroscopy
• Data Analysis

Background:

• Essential information (EI) is crucial for multivariate curve resolution (MCR) and spectral unmixing, representing the most dissimilar data points.
• Current methods for assessing EI rely on principal component analysis (PCA) after all spectral data are acquired, limiting real-time analysis.
• The speed and compression capabilities of EI assessment make it valuable for pre-MCR analysis.

Purpose of the Study:

• To propose a novel approach for extracting essential information in the Fourier domain (EIFD).
• To enable individual and independent assessment of spectral essentiality, overcoming limitations of existing methods.
• To accelerate time-consuming spectral data acquisition processes, such as Raman spectral imaging.

Main Methods:

• Transformed spectral data into Fourier coefficients.
• Assessed EI using convex hull analysis on 2D phasor plots of selected harmonics.
• Utilized a chemometric-driven procedure for targeted data acquisition based on EI content.

Main Results:

• Developed and validated the EIFD method for essential information extraction.
• Demonstrated that EIFD allows for individual and independent evaluation of spectral essentiality.
• Achieved a potential 50-fold acceleration in Raman spectral imaging acquisition through targeted sampling.

Conclusions:

• The EIFD method offers a significant advancement in assessing essential spectral information.
• This approach enables substantial acceleration of spectral data acquisition, particularly for techniques like Raman imaging.
• EIFD facilitates efficient and targeted data collection, improving the speed and practicality of MCR and spectral unmixing.