Rapid spectral analysis for spectral imaging

What is it about?

Describes spectral analysis of images acquired by a hyper-spectral camera. For each pixel, the program converts a reflectance spectrum R(lambda) [dimensionless], lambda is wavelength, into a set of tissue parameters: B = blood content, S = oxygen saturation of blood, a = reduced scattering at 500 nm lambda, M = melanosome volume fraction in epidermis.

Featured Image

Why is it important?

Introduces a rapid matrix inversion approach toward solving for B,S,a,M. The method assumes a tissue water content (W), e.g. W = 0.65, and the program uses the spectrum R(800-1000nm) to specify the scattering strength. In other words, it uses water as an intrinsic calibration standard to specify the scattering. The error in estimating W is low (~±10%) so the error in scattering is low.