Experimental and theoretical comparison of NIR spectroscopy measurements of cerebral hemoglobin changes

Two near-infrared spectroscopy (NIRS) methods are available for measuring changes (Delta) in total cerebral hemoglobin concentration (CHC): 1) a continuous measurement of the changes in total hemoglobin concentration (Delta[Hb](tot)) and 2) the difference between two absolute measurements of CHC, each derived from a small, controlled change in inspired O(2) fraction. This paper investigates the internal consistency of these two methods by using an experimental and theoretical comparison. NIRS was used to measure [Hb](tot) in five newborn piglets before and after a change in arterial PCO(2). Delta[Hb](tot) demonstrated a low coefficient of variation of 2.8 +/- 2.8 (SD) % which allowed changes in CO(2)-cerebral blood volume reactivity to be clearly discriminated. However, a high coefficient of variation of 22.8 +/- 3.5% on the Delta CHC measurements obscured any CO(2) reactivity changes. A theoretical analysis demonstrates the effects of optical pathlength, background absorption, scatter, and blood vessel diameter on both methods. For more accurate monitoring of CHC, individual measurements of optical pathlength and more accurate pulse oximetry are required.