Rapid inverse radiative transfer solver for multiparameter spectrophotometry without integrating sphere.

Area of Science:

• Optical Physics
• Biomedical Optics
• Materials Science

Background:

• Multiparameter spectrophotometry (MPS) is crucial for analyzing turbid materials in diverse fields like clinical diagnosis and food safety.
• Accurate characterization of radiative transfer (RT) parameters is essential for reliable material analysis.

Purpose of the Study:

• To develop and validate a rapid inverse solver using particle swarm optimization (PSO) for retrieving RT parameters (absorption coefficient, scattering coefficient, anisotropy factor) of turbid samples.
• To enhance the efficiency and applicability of MPS for material characterization.

Main Methods:

• Monte Carlo (MC) simulations were employed to generate calculated signals for comparison with measured diffuse reflectance, diffuse transmittance, and forward transmittance.
• A novel objective function was developed and integrated with the PSO algorithm to iteratively refine MC simulations for MPS data.
• The objective function utilizes local averaging of an inverse squared error sum to minimize signal variance in the RT parameter space.

Main Results:

• The developed objective function significantly reduced variance in calculated signals.
• The RT parameters of 20% Intralipid solutions were accurately determined across a wavelength range of 520-1000 nm.
• The entire measurement and inverse calculation process averaged approximately 2.7 minutes per wavelength.

Conclusions:

• The rapid inverse solver facilitates the development of user-friendly and cost-effective MPS instruments.
• This technology enables precise material characterization by resolving compositional profiles at molecular and particulate levels without requiring an integrating sphere.