Objective Magnetic resonance imaging allows the noninvasive observation of PO 2 changes between air breathing and oxygen breathing through quantification of the magnetic longitudinal relaxation time T 1 . Changes in PO 2 are proportional to changes in the longitudinal relaxation rate DR 1 (where DR 1 = 1/ T 1oxygen À 1/T 1air ). Knowledge of this response could inform clinical interventions using maternal oxygen administration antenatally to treat fetal growth restriction. We present in vivo measurements of the response of the fetal-placental unit to maternal hyperoxia.
Design Prospective cohort.Setting Large tertiary maternity hospital.Sample Nine women undergoing low-risk pregnancy (21-33 weeks of gestation) and five nonpregnant adults.Methods During imaging the air supply to mothers was changed from medical air (21% oxygen) to medical oxygen (100% oxygen) and T 1 was monitored over time in both the placenta and fetal brain using a periodically repeated magnetic resonance imaging sequence. To demonstrate that the method could detect a brain response, brain responses from five normal adult volunteers were measured using a similar imaging protocol.Main outcome measure Changes in T 1 following oxygen challenge.Results No significant DR 1 (P = 0.42, paired t-test) was observed in fetal brains. A significant placental DR 1 (P = 0.0002, paired t-test) of 0.02 AE 0.01/s (mean AE SD) was simultaneously observed in the same participants. In the brains of the nonpregnant adults, a significant DR 1 (P = 0.01, paired t-test) of 0.005 AE 0.002/s was observed.Conclusion Short-term maternal oxygen administration does not improve fetal brain oxygenation, in contrast to the response observed in the adult brain.