T2-prepared velocity selective labelling: A novel idea for full-brain mapping of oxygen saturation.

Abstract Background and aim Disturbances in cerebral oxygenation saturation (SO 2 ) have been linked to adverse outcome in adults, children, and neonates. In intensive care, the cerebral SO 2 is increasingly being monitored by Near-InfraRed Spectroscopy (NIRS). Unfortunately NIRS has a limited penetration depth. The “modified T 2 -prepared Blood Imaging of Oxygen Saturation” (T 2 -BIOS) MR sequence provides a step towards full brain SO 2 measurement. Materials and methods Tissue SO 2 , and venous SO 2 (S v O 2 ) were obtained simultaneously by T 2 -BIOS during a respiratory challenge in ten healthy volunteers. These two measures were compared to SO 2 that was obtained by a single probe MR-compatible NIRS setup, and to cerebral blood flow and venous SO 2 that were obtained by arterial spin labelling and T 2 -TRIR, respectively. Results SO 2-T2-BIOS and SO 2-NIRS had a mean bias of − 4.0% (95% CI − 21.3% to 13.3%). S v O 2-T2-BIOS correlated with SO 2-NIRS (R 2  = 0.41, p = 0.002) and S v O 2-T2-TRIR (R 2  = 0.87, p = 0.002). In addition, SO 2-NIRS correlated with S v O 2-T2-TRIR (R 2  = 0.85, p = 0.003) Frontal cerebral blood flow correlated with SO 2-T2-BIOS (R 2  = 0.21, p = 0.04), but was not significant in relation to SO 2-NIRS . Discussion/conclusion Full brain SO 2 assessment by any technique may help validating NIRS and may prove useful in guiding the clinical management of patient populations with cerebral injury following hypoxic-ischaemic events. The agreement between NIRS and T 2 -BIOS provides confidence in measuring cerebral SO 2 by either technique. As it stands now, the T 2 -BIOS represents a novel idea and future work will focus on improvements to make it a reliable tool for SO 2 assessment.