Optimizing a two-layer method for hybrid diffuse correlation spectroscopy and frequency-domain diffuse optical...

Area of Science:

• Biomedical Optics
• Neuroimaging
• Medical Physics

Background:

• Diffuse optics techniques are susceptible to confounding signals from extracerebral tissues.
• Two-layer (2L) head models offer a way to distinguish cerebral signals from extracerebral artifacts.
• Conventional 2L models may suffer from crosstalk between fitting parameters, affecting accuracy.

Purpose of the Study:

• To implement and validate a constrained 2L head model for combined diffuse correlation spectroscopy (DCS) and frequency-domain diffuse optical spectroscopy (FD-DOS) data.
• To quantify the accuracy of the proposed model in estimating cerebral blood flow and tissue absorption.
• To assess the model's robustness against noise and parameter crosstalk.

Main Methods:

• Utilized an analytical solution for a 2L cylinder model with a priori extracerebral layer thickness.
• Fitted multi-distance FD-DOS and DCS data, assuming homogeneous tissue scattering.
• Validated the algorithm using simulated data (slab and realistic head geometries) and in vitro phantom experiments.

Main Results:

• The constrained 2L algorithm achieved low errors in cerebral flow index (6.3% for slab, 34% for head geometries).
• Cerebral absorption coefficient errors were 5.0% (slab), 4.6% (head), and 8% (phantom).
• Results demonstrated minimal sensitivity to scattering variations and robustness to parameter crosstalk.

Conclusions:

• The constrained 2L algorithm effectively separates cerebral signals and reduces artifacts in diffuse optics measurements.
• This model shows improved accuracy for cerebral blood flow and absorption estimation compared to semi-infinite models.
• The approach holds promise for enhancing the reliability of FD-DOS and DCS in adult studies.