Optical Monitoring of Cerebral Hemodynamics and Oxygenation in the Neonatal Piglet

> Objective: Optical spectroscopy can monitor changes in cerebral oxyhemoglobin (HbO), deoxyhemoglobin (Hb), and total hemoglobin (HbT = Hb + HbO). The purpose of this study was to identify which one of these three optical variables would be the best correlate of cerebral hemodynamics and oxygenation accompanying different physiological conditions known to affect cerebral perfusion pressure and oxygenation. Methods: Eleven anesthetized and ventilated newborn piglets were subjected to: 1) hyperoxic hypercapnia, 2) asphyxia from room air, 3) hypertension, 4) hypotension, and 5) cerebral venous congestion. Optical measurements were recorded continuously along with carotid blood flow (CaBF) measured by the transit time Doppler technique and arterial hemoglobin oxygen saturation (SaO2) measured by pulse oximetry. Results: A good positive HbT-CaBF correlation was recorded during hypercapnia (r = 0.816), asphyxia (r = 0.846), hypertension (r = 0.72), and hypotension (r = 0.95), where as negative correlation was recorded during venous congestion (r = -0.91). A good positive correlation HbO-SaO2 was recorded during asphyxia (r = 0.842) and hypotension (r = 0.86), whereas no HbO-SaO2 agreement occurred during hypercapnia (r = 0.23), hypertension (r = 0.19), and venous congestion (r = 0.34), when SaO2 did not change at all. A good positive HbO-SaO2 correlation was recorded during hypercapnia (r = 0.84), hypertension (r = 0.91), and hypotension (r = 0.94), suggesting a high HbO dependence on changes in CaBF. Conclusions: We demonstrated that none of the optical variables (Hb, HbO, HbT) alone can be used as a correlate of cerebral hemodynamics and oxygenation over a broad range of cerebral perfusion pressure and oxygenation. However, all three variables combined (i.e. optical patterns) strongly reflect the changes in cerebral hemodynamics and oxygenation accompanying those physiological conditions. These and similar patterns could potentially serve as a reference model for future studies in the human fetus where "gold standard" correlation methods do not exist.