Toward routine assessment of cerebral blood flow in neonates and infants: a phase-contrast MRI study
Peiying Liu1, Ying Qi2, Zixuan Lin1, Xuna Zhao3, Qiyong Guo2, Xiaoming Wang2, and Hanzhang Lu1

1Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, 2Shengjing Hospital of China Medical University, Shenyang, China, People's Republic of, 3Philips Healthcare, Beijing, China, People's Republic of

Synopsis

Knowledge of CBF in neonates or infants may provide valuable information in many pathological conditions. When applied to very young children, CBF mapping using arterial-spin-labeling (ASL) MRI suffers from low SNR and poor quantification, whereas phase-contrast (PC) MRI may provide reliable estimation of global CBF. Therefore, this study aim to 1) provide a set of age-specific PC-MRI protocols for CBF quantification in children under 1.5 years old; 2) establish typical arterial flow velocity in children at this age which could guide future ASL efforts in labeling pulse optimization; 3) report how CBF changes during this early stage of life.

Purpose

CBF is an important indicator of brain function and tissue viability. Knowledge of CBF in neonates or infants may provide valuable information in pathological conditions such as hypoxic ischemic injury1, perinatal ischemic stoke2, and congenital heart disease3. However, routine clinical assessment of neonatal or infant CBF is difficult due to a lack of techniques. There is a recent surge in using arterial spin labeling (ASL) MRI for neonatal CBF assessment1,2. However, current ASL methods when applied to very young children suffer from both low SNR and poor quantification. Furthermore, neonatal/infant stage is characterized by rapid changes in brain physiology and it is virtually impossible to find a one-size-fits-all imaging protocol. In this regard, phase-contrast (PC) MRI is complementary to ASL in that its pulse sequence is relatively simple and it provides a reliable and quantitative, although lack spatial information, estimation of CBF3-5. Therefore, the goal of the present study is to 1) provide a set of age-specific protocols for CBF quantification in children from 34 gestational weeks to 1.5 years after birth; 2) establish typical arterial flow velocity in children at this age which could guide future ASL efforts in terms of optimal labeling pulse design; 3) report how CBF changes during this early stage of life. We also recommend that PC-MRI be used concomitantly with ASL to provide a normalization of ASL-derived CBF maps to ensure quantification accuracy6.

Methods

MRI data were collected on a 3T Philips system and the data were used upon institutional ethics committee approval. A total of 24 children were studied.

Visualization of feeding arteries with a minimal scan duration

Since the positioning of PC-MRI requires the visualization of the brain’s feeling arteries, we first optimized the TOF angiogram in four infants. The goal was to identify an optimal spatial resolution to obtain a tradeoff between scan time and artery delineation. We compared three imaging resolutions: 0.6mm, 0.8mm and 1mm. Other parameters were similar to a previous study4. The scan duration was 20s, 24s, and 32s for the three resolutions, respectively.

Optimization of cut-off velocity (Venc) in PC-MRI

PC-MRI uses a pair of magnetic field gradients to encode the flow velocity of a spin in its phase. The most important imaging parameter in PC-MRI is the Venc. Twenty infants (34~114 gestational weeks) were included. In each infant, we performed PC-MRI 12 times: six times on the left internal carotid artery (LICA) using a series of Venc values and six times on the left vertebral artery (LVA). Venc for LICA were 10 to 60cm/s with 10cm/s increment, whereas Venc for LVA were 5 to 30cm/s with 5cm/s increment. Other imaging parameters were: FoV=90x90x3mm3, matrix size=180x180. For each PC scan, we evaluated the blood flux and peak velocity of the targeted artery.

Age dependence of CBF

The total blood flux to the brain was estimated as (fluxLICA+fluxLVA)x2. Brain volume was obtained from T2-weighted anatomic scans. Unit-volume CBF in ml/100g/min was calculated. Age-changes in total arterial flux, brain volume and CBF was assessed.

Results and Discussion

Figure 1 shows TOF images at three different resolutions in a neonate at 37 gestational weeks. Consensus among three investigators by inspecting data from all four subjects suggested that in-plane resolution of 0.8mm provides an optimal tradeoff between image quality and scan duration. This TOF protocol was used in the rest of the subjects.

Figure 2 shows the PC-MRI results from one representative subject (35 gestational weeks). Figure 3 summarizes the peak velocity across all subjects. Both ICA and VA showed a gradual increase in peak velocity. The optimal Venc of each participant is shown in red dots in Figure 3, and fitting these dots to a step-wise curve yielded the age-specific Venc recommendations in the green curve.

Figure 4 shows age-related changes in total arterial flux (in ml/min), brain volume (in ml), and unit-volume CBF (in ml/100g/min). All three parameters increased with age. For unit-volume CBF, it appears that, at birth, the neonate’s CBF is only one third of that of adult (typically 60 ml/100g/min). But it rapidly increases with time and, by six months after birth, CBF is already at the level of adults. Beyond that age, the CBF continues to increase to a level greater than adults. According to previous literature7, CBF is anticipated to peak between 3 to 8 years old, and then started to decrease thereafter.

Conclusion

Here we propose a procedure, shown in Figure 5, for the quantitative assessment of CBF in children under 1.5 years of age. Our results also demonstrated a rapid increase of CBF during this period.

Acknowledgements

None

References

1. Pienaar et al., NeuroImage, 63:1510, 2012; 2. De Vis et al., Pediatr Res 74:307, 2013; 3. Jain et al., JCBFM 34:380, 2014; 4. Varela et al., NMR Biomed 25:1063, 2012; 5. Liu et al., NMR Biomed 27:332, 2014; 6. Aslan et al, MRM 63:765, 2010; 7. Takahashi et al., AJNR 20:917, 1999.

Figures

Figure 1. The TOF angiogram MIP images of one subject (37.1 gestational weeks). The 0.8mm scan allows a better visualization of the small arteries than the 1mm scan, while taking less time compared to the 0.6mm scan.

Figure 2. Positioning of the PC-MRI scans and the results from one subject (35.4 gestational weeks). (a) PC-MRI scan positioning for LICA and LVA. (b) Peak velocity in the targeted artery measured at different Venc of PC-MRI.

Figure 3. Summary of peak velocity and recommended Venc across all subjects. (a) ICA results. (b) VA results. Green lines indicate the recommended optimal Venc for PC-MRI at specific age.

Figure 4. Blood flux, brain volume, and CBF as a function of age. (a) Age effect on total flux to the brain. (b) Age effect on brain volume. (c) Age effect on unit-volume CBF.

Figure 5. Proposed MRI procedure for a complete CBF dataset in children under 1.5 years of age.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
0559