CC BY
0Separate reconstruction of absorption and scattering coefficients spectra from the diffuse reflectance spectroscopy data based on the refined analytical model
E. Sergeeva1, A. Kostyuk1, D. Kurakina1, A. Khilov1, M. Kirillin1
1-A.V. Gaponov-Grekhov Institute of Applied Physics RAS; 46 Ulyanov st., Nizhny Novgorod 603950 Russia
* sergeeva_ea@ipfran.ru
Despite its technical simplicity and decades of practical implementation in experimental biophotonics, diffuse reflectance spectroscopy (DRS) is still an in-demand technique for fast noninvasive quantification of biological chromophores and tracking dynamical changes in their concentrations in various biotissues including studies of brain hemodynamics and superficial biotissues. Systems equipped with multiple detectors such as hyperspectral cameras or multifiber setups are aimed at mapping the chromophores distribution from the absorption coefficient spectra reconstructed from the measured spectra of diffuse reflectance at various source-detector distances (SDDs). There are numerous approaches to solve the DRS inverse task: from approximate analytical models involving calibrated reflectors measurements and up to machine learning algorithms based on numerical solutions of a forward task in a realistic geometry. For large volumes of processed data a time-saving recovery algorithm based on an adequate analytical model is preferrable. The most common approach to recover the chromophores content from DRS data is to fit the reflectance spectra or the ratio of spectra measured at one or several SDDs by the analytically calculated reflectance where the absorption spectrum is assumed as a weighted sum of partial spectra of a priori known chromophores, and the reduced scattering spectrum is fitted by a given function with a small number of parameters [1]. However, in some cases the basic set of chromophores may be incomplete so the composed guess spectrum does not reflect the features of real absorption spectrum. Besides, in biotissues with spatially non-uniform distribution of chromophores such as skin the sensitivity of a wideband DRS system to different layers varies for different parts of the wavelength range, which violates the assumption of constant partial weights of basic chromophores.
We propose an algorithm for separate, condition-free reconstruction of absorption and reduced scattering spectra of biotissue from DRS data for small SDDs (below 5 mm) based on the refined analytical model [2]. The algorithm has been verified by Monte Carlo simulated data mimicking wideband DRS spectra measured from uniform and two-layered skin structures that emulate dermis and epidermis-dermis geometry, respectively. Further the algorithm has been tested in the model and in vivo experiments in volunteers performed with the custom-made DRS system operating in the wavelength range of 450-1000 nm [3]. We discuss the accuracy of the algorithm in separate reconstruction of absorption and scattering spectra of a liquid phantom of biotissue which optical properties and preliminary evaluated from the spectrophotometry measurements. We demonstrate examples of determination of optical properties of skin in different locations in vivo and discuss the ability and limitations to recover the chromophore composition.
The study has been supported by Center of Excellence "Center of Photonics" funded by The Ministry of Science and Higher Education of the Russian Federation, contract № 075-15-2022-316.
[1] S.L. Jacques, Optical properties of biological tissues: a review, Phys. Med. Biol., vol.58, pp. R37-R61 (2013).
[2] E. Sergeeva. D. Kurakina, I. Turchin, M. Kirillin, A refined analytical model for reconstruction problems in diffuse reflectance spectroscopy, Journal of Innovative Optical Health Sciences, 2342002 (2023).
[3] V. Perekatova, A. Kostyuk, M. Kirillin, E. Sergeeva, D. Kurakina, O. Shemagina, A. Orlova, A. Khilov, I. Turchin, VIS-NIR diffuse reflectance spectroscopy system with Self-calibrating fiber-optic probe: study of perturbation resistance, Diagnostics (Basel), vol.13, pp. 457 (1-20) (2023).
