MRI-based blood oxygen saturation measurements in infants and children with congenital heart disease

Abstract

Background: Vessel oxygen saturation can be determined with MR oximetry using an in vivo measurement of signal decay (T2) and the calibration curve relating T2 to blood oxygen saturation (%HbO₂), where: 1/T2=1/T2O+K (1–%HbO₂/100)² and K is a constant parameter which correlates with measured fibrinogen levels. The ability to noninvasively measure %HbO₂ in cardiac chambers and vessels has enormous potential in children with congenital heart disease (CHD). Objective: The purpose of the study was to prospectively characterize the T2–%HbO₂ relationship in infants where T2–%HbO₂ is the realtionship between T2 and %HbO₂ (blood oxygen saturation) expressed by the equation given above, and to determine whether adult values for K and T2O (where T2O is the T2 of fully oxygenated blood) can be extrapolated to pediatric patients with CHD. A second objective was to apply this method to calculate the %HbO₂ in vivo using MR imaging in infants with CHD. Materials and methods: Fifteen patients with CHD undergoing cardiac catheterization (9 male; 6 female), median age 8 months, were recruited for the calibration study. T2O and K were measured directly from blood samples, compared with the values estimated from adult population statistics, and plotted against hematocrit and fibrinogen, respectively. In four studies of infants with CHD, T2 measurements were converted to %HbO₂ using the calibration curve. Results: The T2–%HbO₂ relationship in infants correlated with the adult calibration statistics (1/T2O vs. hematocrit, r=0.77; K vs. fibrinogen, r=0.61). Our initial in vivo studies demonstrated that the MR oximetry reflected the expected oxygen saturations. Conclusion: Adult values for T2–%HbO₂ calibration can be used to measure blood oxygen saturation in vivo in children with CHD.