Using T₂-relaxation under spin tagging (TRUST)^1–3 to measure cerebral oxygenation in neonates has shown that subject motion affects blood signal isolation and has unknown impact on venous oxygen saturation (SvO2) quantification. In TRUST, the venous blood signal in the superior sagittal sinus (SSS) is isolated using a spin tagging approach, an exponential fit is used to estimate T₂ from the resulting signal, and an empirical calibration is used to map T₂ to venous oxygen saturation. To quantify subject motion during TRUST scans, we incorporated volume navigators (vNav)⁴ into the TRUST sequence to retrospectively determine the movements that occurred during the scan. We show that for static scans vNav modules have only small effects on resulting venous blood T₂ estimates, that poor exponential goodness of fit is not a sufficient indicator of motion, and that blood T₂ is biased upwards with increasing motion.vNav acquisition modules⁴ were inserted into the TRUST sequence (Figure 1). For offline registration not including the label shadow we used the methods of McDaniel, et al:⁵ volumes were co-registered and motion parameters measured using the FSL FLIRT tool (cost function=correlation ratio, DOF=6)⁶ and custom Matlab scripts. Translations were determined for a voxel centered on the SSS and motion score calculated.⁴ For this pilot study with compliant subjects, three adults (2 male, 1 female, mean age=21) were scanned with IRB approval on a Siemens Tim Trio scanner. The TRUST image was positioned 25 mm above the confluence of the sinuses perpendicular to the SSS with scan parameters: TE/TR=12/5000ms, T2 effective echo times (TE_eff)=0,18,36,72,144ms, resolution=3.4x3.4x5mm, inversion time=1200ms, tagging width=100mm, tagging gap=25mm, 3 averages, total acquisition=2:30. The subject was then instructed to move or remain still for alternating TRUST+vNav scans. The three brightest voxels near the SSS were averaged to give a signal intensity (S(TE_eff)) value for one TE_eff. Fitting to S(TE_eff)=S₀exp(TE_eff*C) was performed by taking ln(S(TE_eff)) and then performing a linear least squares fit. T_2,blood=1/(R_1,blood-C), R_1,blood=0.62 [s^-1].¹ Goodness of fit was evaluated based on the standard deviation of the residuals (SDR) of the fit.Simulations of the TRUST+vNav sequence using Bloch simulation and TRUST signal equations¹ show that vNav modules cause changes in T₂ smaller than same subject T₂ variance and SDR changes were smaller than for typical in vivo fits. T₂ estimates from TRUST and TRUST+vNav sequences demonstrate no significant difference in T₂ values (Table 1). vNavs track head motion, with example motion trajectories for a quiescent and moving subject shown in Figure 2. Figure 3 shows maximum motion score during the scan plotted against SDR. Figure 4 shows the significant positive correlation between the difference of estimated T₂ and mean baseline T₂ per subject, and maximum motion score.vNavs have been well characterized and are suitable for resolving translations of less than 0.5mm.⁴ Due to the differences between control and label acquisitions in TRUST, a small (<0.5 mm translation at the SSS) jitter results from the registration between these states for still subjects. Figure 2 shows the registration across all label acquisitions, and future work will consider the motion during control and label states in parallel since the venous blood signal isolation is affected by either. Figure 2 suggests good motion tracking using the vNavs. Previous work established that the failed signal isolation resulting from movement between control and tag acquisitions partly manifests itself as poor goodness of exponential fit.¹ Figure 3 suggests that though this is largely true, goodness of fit as assessed by SDR cannot be used to determine whether or not motion occurred. This is important both because of the upward bias in T₂ estimates that results from motion (Figure 4) and the estimation error in T₂ that grows larger with increased motion. We hypothesize that the larger T₂ values with motion result from poor venous blood signal isolation, meaning that tissue compartments with higher intrinsic T₂ values (grey matter and cerebral spinal fluid) are included in the assumed pure blood signal. Future work will use the motion parameters—measured independently from the T₂-fitting process—to perform retrospective motion correction by rejecting or weighting S(TE_eff) values before performing the fit. We will also explore the possibility that brain state changes contribute to the upward bias in T2 when a subject is instructed to move.Motion causes an upward bias of venous blood T₂ estimates in the TRUST technique. During scans with poor subject compliance, specifically during pediatric scans, independent motion monitoring, here performed with vNavs, may be necessary.