Cerebral metabolic rate of oxygen (CMRO2) is an important biomarker for normal and pathological neonatal brain development. However, regional measurement of neonatal CMRO2 has been limited due to an inability to evaluate regional venous blood oxygenation (Yv). This study presented a rapid MRI technique, accelerated T2-relaxation-under-phase-contrast (aTRUPC), to measure vessel-specific Yv in neonates. We have improved the reliability and accuracy of aTRUPC by optimizing imaging parameters and calibrating T2-bias. A pilot study on healthy, non-sedated neonates demonstrated the feasibility of aTRUPC to measure vessel-specific Yv. Accuracy of aTRUPC-based Yv measurements was further validated with established whole-brain Yv measurement using T2-relaxation-under-spin-tagging.
Pulse sequence: The proposed pulse sequence, accelerated T2-relaxation-under-phase-contrast (aTRUPC), was a substantially accelerated implementation of the original TRUPC sequence.4 As seen in Figure 1, the bipolar gradients in phase-contrast modules isolate pure blood signal from static tissues, and the T2-preparation allows the quantification of blood T2 which can be converted to Yv via T2-Yv calibration plots.7,8 Drastic reduction in scan time was achieved by exploiting the turbo-field-echo (TFE) scheme (i.e., applying a train of phase-contrast acquisition modules after each T2-preparation). To minimize potential bias in T2-estimation and image blurring, centric k-space sampling9 and variable flip-angle (FA) scheme10 were implemented. The FAs were designed to keep the blood signal constant throughout the acquisition train.
Optimization of aTRUPC protocol: Optimization studies were performed in 8 healthy, non-sedated neonates (5 females, gestational age 39.8±0.6weeks) on a 3T Siemens Skyra system. Three TFE factors (5, 10 and 15), four encoding-velocities (VENCs) (3, 5, 7, 9cm/s), and two slice thicknesses (5 and 10mm) were compared using eTE=0ms scans, respectively. Signal-to-noise ratio (SNR), reproducibility and scan duration were considered in the selection of optimal parameters.
Calibration of T2-estimation bias: Numerical Bloch simulations showed systematic T2-estimation bias in aTRUPC associated with the TFE train, which is dependent on actual T2 but not T1 (details not shown due to space limit). Therefore, an in vivo study on adults was performed to establish a T2-bias calibration curve for the correction of this bias. In five healthy adults (26.6±3.9y) scanned on a 3T Siemens Prisma system, we compared T2 values from 7 ROIs measured by aTRUPC using TFE=15 with those measured by original TRUPC which has no TFE-induced T2-bias. Linear regression between the two measurements was used to establish the T2-bias calibration curve.
Feasibility in neonates: The optimized aTRUPC was then performed in four healthy, non-sedated neonates (2 females, gestational age 40.0±0.5weeks). To enhance SNR, 4 averages were acquired. The T2 estimated by aTRUPC was corrected using the T2-bias calibration curve before conversion to Yv. Yv map and 8 ROI values were obtained. For validation, the Yv in posterior superior-sagittal-sinus (SSS) measured by aTRUPC was compared to the established global Yv measurement using T2-relaxation-under-spin-tagging (TRUST) MRI.11
The comparisons among different TFE factors are shown in Figure 2. It was found that a TFE factor of 15 has a higher SNR than with a TFE factor of 5 (P=0.04) or 10 (P=0.03). A linear mixed effect model analysis revealed that a higher TFE factor resulted in higher reproducibility (i.e., lower CoV) (P=0.004), probably due to shorter scan duration and thus less motion. Therefore, we concluded that a TFE factor of 15 was optimal. Using similar analyses, we determined an optimal slice thickness of 10mm and an optimal VENC of 5cm/s. The scan duration of the optimized aTRUPC sequence (with 3 eTEs) was 56s, which was nearly 1/5 of the original TRUPC’s duration (4.8min).
Figure 3 shows the results of the T2-bias calibration. A strong linear correlation was found between the T2 measurements by aTRUPC and the original TRUPC. Based on the regression analysis, the T2-bias calibration curve was determined to be: $$$T_{2,corrected}=1.1397{\times}T_{2,aTRUPC}-7.1143$$$
Figure 4a displays the Yv maps of all 4 neonates and Figure 4b showed the scatter plot of their posterior SSS Yv using TRUST and aTRUPC. A good agreement between the two methods can be observed, reflecting the accuracy of aTRUPC-measured Yv. Figure 4c illustrates the group-averaged Yv results. Vessels draining deep brain tend to have higher Yv than those draining the cortex, which agrees with previous adult literature results.4
1. Douglas-Escobar M, Weiss MD. Hypoxic-ischemic encephalopathy: a review for the clinician. JAMA Pediatr 2015;169:397-403.
2. van der Aa NE, Benders MJ, Groenendaal F, de Vries LS. Neonatal stroke: a review of the current evidence on epidemiology, pathogenesis, diagnostics and therapeutic options. Acta Paediatr 2014;103:356-364.
3. Liu P, Chalak LF, Lu H. Non-invasive assessment of neonatal brain oxygen metabolism: A review of newly available techniques. Early Hum Dev 2014;90:695-701.
4. Krishnamurthy LC, Liu P, Ge Y, Lu H. Vessel-specific quantification of blood oxygenation with T2-relaxation-under-phase-contrast MRI. Magn Reson Med 2014;71:978-989.
5. Bolar DS, Rosen BR, Sorensen AG, Adalsteinsson E. QUantitative Imaging of eXtraction of oxygen and TIssue consumption (QUIXOTIC) using venular-targeted velocity-selective spin labeling. Magn Reson Med 2011;66:1550-1562.
6. Guo J, Wong EC. Venous oxygenation mapping using velocity-selective excitation and arterial nulling. Magn Reson Med 2012;68:1458-1471.
7. Liu P, Chalak LF, Krishnamurthy LC, Mir I, Peng SL, Huang H, Lu H. T1 and T2 values of human neonatal blood at 3 Tesla: Dependence on hematocrit, oxygenation, and temperature. Magn Reson Med 2016;75:1730-1735.
8. Lu H, Xu F, Grgac K, Liu P, Qin Q, van Zijl P. Calibration and validation of TRUST MRI for the estimation of cerebral blood oxygenation. Magn Reson Med 2012;67:42-49.
9. Ding H, Fernandez-de-Manuel L, Schar M, Schuleri KH, Halperin H, He L, Zviman MM, Beinart R, Herzka DA. Three-dimensional whole-heart T2 mapping at 3T. Magn Reson Med 2015;74:803-816.
10. Busse RF, Brau AC, Vu A, Michelich CR, Bayram E, Kijowski R, Reeder SB, Rowley HA. Effects of refocusing flip angle modulation and view ordering in 3D fast spin echo. Magn Reson Med 2008;60:640-649.
11. Liu P, Huang H, Rollins N, Chalak LF, Jeon T, Halovanic C, Lu H. Quantitative assessment of global cerebral metabolic rate of oxygen (CMRO2) in neonates using MRI. NMR Biomed 2014;27:332-340.