Improved accuracy of brain oxygen metabolism measurements using multi-distance diffuse correlation spectroscopy and near infrared spectroscopy together with a Monte Carlo light transport model (Conference Presentation)

Abstract

Diffuse correlation spectroscopy (DCS) is being employed alongside near-infrared spectroscopy (NIRS) measurements to track the cerebral oxygen metabolic rate (CMRO2). However, both techniques employ diffusely reflected light that has traveled mostly through extracerebral tissues. Recent studies indicate that depth sensitivity profiles are different for NIRS vs DCS measurements, with DCS appearing to be more sensitive to the brain than NIRS methods for a given source-detector separation. This mismatch can lead to erroneous conclusions with respect to the amount and perhaps even the direction of change in CMRO2. Recently, our group and others have demonstrated the use of Monte Carlo (MC) based multi-layer, multi-distance fitting, which offers increased accuracy for complex tissue structures such as the adult brain. In this paper we employ a Monte Carlo light transport model based on a realistic head geometry that can be derived from MRI scans (if available) or approximated from head shape measurements. We consider DCS and CW-NIRS measurements taken at two or more distances and analyze simulated data generated using a fully segmented adult brain MRI scan. Through simulations, we explore the improvements offered by our method vs. processing the same measurements with a semi-infinite diffusion model and estimate the impact of errors in geometry and optical properties on relative blood flow and CMRO2 changes.