Accuracy of diffuse correlation spectroscopy measurements of cerebral blood flow when using a three-layer analytical model

Abstract

Diffuse correlation spectroscopy (DCS) is a non-invasive optical technology for the assessment of an index of cerebral blood flow (CBFi). Analytical methods that model the head as a three-layered medium (i.e., scalp, skull, brain) are becoming more commonly used to minimize the contribution of extracerebral layers to the measured DCS signal in adult cerebral blood flow studies. However, these models rely on a priori knowledge of layer optical properties and thicknesses. Errors in these values can lead to errors in the estimation of CBFi, although the magnitude of this influence has not been rigorously characterized. Herein, we investigate the accuracy of measuring cerebral blood flow with a three-layer model when errors in layer optical properties or thicknesses are present. Through a series of in silico experiments, we demonstrate that CBFi is highly sensitive to errors in brain optical properties and skull and scalp thicknesses. Relative changes in CBFi are less sensitive to optical properties but are influenced by errors in layer thickness. Thus, when using the three-layer model, accurate estimation of scalp and skull thickness are required for reliable results.

Influence of source–detector separation on diffuse correlation spectroscopy measurements of cerebral blood flow with a multilayered analytical model

“…Data were simulated using the three-layer solution to the correlation diffusion equation for the normalized intensity autocorrelation function, , at delay time, , and SDS, , as in Refs. 12 and 16 . To investigate the influence of , data were simulated at 0.5, 1.0, 1.5, 2.0, 2.5, and 3 cm.…”

“…For each , we simulated at a range of optical properties and skull/scalp thicknesses ( , , , , , , , and ). 16 For each parameter, we sampled 11 evenly spaced values within the reported range while fixing the other parameters at the median of the corresponding range. For each of these 11 values, we simulated 12 evenly spaced values of CBFi and six values of scalp blood flow index (SBFi) .…”

“…For each r, we simulated g 2 ðr; τÞ at a range of optical properties and skull/scalp thicknesses (μ a;scalp ∈ ½0.05; 0.15 cm −1 , μ a;skull ∈ ½0.05; 0.15 cm −1 , μ a;brain ∈ ½0.05; 0.25 cm −1 , μ 0 s;scalp ∈ ½8;12 cm −1 , μ 0 s;skull ∈ ½8;12 cm −1 , μ 0 s;brain ∈ ½2;6 cm −1 , L scalp ∈ ½0.15; 0.53 cm, and L skull ∈ ½0.35; 1.10 cm). 16 For each parameter, we sampled 11 evenly spaced values within the reported range while fixing the other parameters at the median of the corresponding range. For each of these 11 values, we simulated 12 evenly spaced values of CBFi ∈ ½2 × 10 −8 ; 9 × 10 −8 cm 2 ∕s and six values of scalp blood flow index (SBFi) ∈ ½1∕8;1∕3 × CBFi.…”

“…Although the importance and accuracy of the three-layer slab analytical model have been investigated in former studies, 12 , 14 , 16 the approach is highly sensitive to measurement noise. Thus, it is important to balance the depth penetration with the signal-to-noise ratio (SNR) when selecting source–detector separation(s) (SDS) to optimize the accuracy of cerebral blood flow estimation using this approach.…”

“…[12][13][14] Recent validation with transcranial Doppler ultrasound (TCD) supports the superiority of the three-layer models over two-layer models and highlights the need for including a negligible flow layer (skull) when assessing cerebral perfusion with DCS. 15 Although the importance and accuracy of the three-layer slab analytical model have been investigated in former studies, 12,14,16 the approach is highly sensitive to measurement noise. Thus, it is important to balance the depth penetration with the signal-to-noise ratio (SNR) when selecting source-detector separation(s) (SDS) to optimize the accuracy of cerebral blood flow estimation using this approach.…”