Physics-based multiplicative scatter correction approaches for improving the performance of calibration models.

Area of Science:

• Spectroscopy
• Chemometrics
• Biophysics

Background:

• Light scattering significantly complicates chemical property analysis in particulate systems like blood and tissues.
• The extended multiplicative signal correction (EMSC) method has been used to empirically account for light scattering.
• Integrating physics-based light scattering models into EMSC offers potential for improved separation of spectral effects.

Purpose of the Study:

• To develop a novel preconditioning method for enhancing the extended multiplicative signal correction (EMSC) in spectral analysis.
• To improve the separation of absorption and scattering effects in particulate systems.
• To create more accurate calibration models for chemical property estimation.

Main Methods:

• A preconditioning step involving a physics-based light scattering transformation was applied to spectra.
• This transformation converts spectra into a form suitable for separable absorption and scattering analysis.
• The transformed spectra were then processed using the extended multiplicative signal correction (EMSC) method.

Main Results:

• The proposed preconditioning step, when combined with EMSC, yielded superior calibration models compared to direct EMSC application.
• The transformation effectively separated absorption and scattering effects, addressing a key limitation of previous methods.
• Improved accuracy in estimating chemical properties of particulate systems was demonstrated.

Conclusions:

• A physics-based spectral transformation followed by EMSC offers a significant advancement for analyzing particulate systems.
• This approach enhances the reliability of chemical property estimation by effectively managing light scattering.
• The method holds promise for applications in fields such as biomedical analysis and pharmaceutical quality control.