Time domain NIRS sensitivity to neonatal brain haemodynamics

Impairment of cerebral autoregulation and haemodynamics is an important cause of morbidity and mortality in neonates, making a continuous monitoring of brain health extremely important in Neonatal Intensive Care Unit. Time Domain NIRS (TD-NIRS) is an advantageous technique for non-invasively continuous monitoring of brain hemoglobin concentration at bed side. However, NIRS measurements are influenced by superficial tissues (i.e., scalp, skull and cerebrospinal fluid) and head geometry, which could affect the estimation of brain haemodynamic parameters. In this work we evaluated the accuracy of TD-NIRS in estimating neonatal cerebral haemodynamics through Monte Carlo simulations. In particular, we simulated photon propagation inside two realistic meshes of preterm (29 weeks post-menstrual age (PMA)), and term (44 weeks PMA) neonates' head. The obtained results were analyzed with semi-infinite and spherical homogeneous model for photon migration, and showed an underestimation of hemoglobin concentration, and good accuracy for brain saturation, empowering the ability of TD-NIRS technique in estimating brain tissue saturation.