New closed-form approximation for skin chromophore mapping.

Area of Science:

• Biomedical Optics
• Photonic Imaging
• Dermatology

Background:

• Estimating skin chromophore concentrations (blood, melanin) is crucial for diagnostics.
• Existing methods like Monte Carlo simulations and optimization-based models are computationally intensive.
• The differential modified Beer-Lambert law offers speed but often lacks accuracy.

Purpose of the Study:

• To develop a computationally efficient and accurate method for skin chromophore mapping.
• To improve the speed of estimating blood and melanin concentrations in spectral images.

Main Methods:

• Derived optimal coefficients for the differential Beer-Lambert model by differentiating a diffusion model.
• Validated the method computationally using Monte Carlo simulations and experimentally with in vivo measurements (Allen's test).
• Applied the model to predict chromophore concentration differences from absorption spectra differences.

Main Results:

• Achieved high correlations for predicted vs. actual blood (r=0.94), melanin (r=0.99), and oxygen saturation (r=0.73).
• Demonstrated a significant speed improvement: ~20 minutes for a 1-megapixel image, orders of magnitude faster than optimization methods.
• The method accurately predicts chromophore concentrations.

Conclusions:

• The proposed method offers a fast and accurate solution for skin chromophore mapping.
• This advancement has potential applications in non-invasive skin diagnostics and imaging.
• The optimized differential Beer-Lambert model provides a viable alternative to slower, less accurate methods.