Whole-Brain CBF and BAT Template Measured by Multi-TI Arterial Spin-Labeling Technique and Its Application in Cerebellar Infarction
Yelong Shen1, Bin Zhao1, Guangbin Wang1, Shuang Yang1, Shan Li2, Josef Pfeuffer3, and Tianyi Qian4

1Shandong Medical Imaging Research Institute, School of Medicine, Jinan, China, People's Republic of, 2Department of Neurology, Provincial Hospital Affiliated to Shandong University, Jinan, China, People's Republic of, 3Siemens Healthcare, Application Development, Erlangen, Germany, Erlangen, Germany, 4Siemens Healthcare, MR Collaborations NE Asia, Beijing, China, People's Republic of

Synopsis

Single-TI ASL usually underestimates the cerebral blood flow in areas with longer blood arrival time, especially in the cerebellum. In this study, we built a template of whole-brain cerebral blood flow and blood arrival time based on multi-inversion time-ASL (mTI-ASL). No significant differences were found when comparing young vs. old groups and female vs. male groups. The application in cerebellar infarction patients demonstrates that mTI-ASL performs better than sTI-ASL especially in areas with longer BAT. The CBF/BAT template created based on normal subjects could be used to better identify perfusion deficits.

PURPOSE:

Cerebral blood perfusion can be measured by dynamic susceptibility contrast (DSC) or arterial spin labeling (ASL). However, it is often difficult to visualize bilateral low CBF using DSC, since it only provides semi-quantitative values, and physician usually rely on the evaluation of interhemispheric differences to identify abnormal perfusion patterns. On the other hand, single-TI ASL tends to underestimates the CBF in areas with longer blood arrival time; especially in the cerebellum. In this study, we built a template of whole-brain cerebral blood flow and bolus arrival time based on multi-inversion time-ASL (mTI-ASL). The BAT/CBF template provides a standard for diagnosing ischemia, and the diagnostic of patients with cerebellar infarction or bilateral ischemic stroke could be facilitated by comparing the individual perfusion maps to these templates.

METHODS:

49 healthy adults (25 males, age ranging from 25 to 50 with median of 36) and 8 patients with cerebellar infarction participated in this study. The MRI exam consisted of Multi-TI ASL, T1 MPRAGE, T2-FLAIR and DWI. All data was collected on a MAGNETOM Skyra 3T MR scanner (Siemens Healthcare, Erlangen, Germany). mTI-ASL images were acquired with a prototype sequence with the following parameters: TR/TE = 4600/22ms, FOV = 220 × 220mm2, slice thickness = 4mm, voxel size = 3.4 x 3.4 × 4.0mm3, 36 slices, bolus length = 700ms, 16 TIs from 480 to 4080ms, and total acquisition time = 5:09 min including an M0 scan. The CBF/BAT/error maps were calculated inline on the scanner. MPRAGE was acquired in sagittal orientation with 1 mm isotropic resolution (FOV = 230 × 230 mm, 192 slices, TR/TE = 1900ms/2.58ms, TI = 900ms, flip angle = 9°) in 4:59 min. Voxels’ CBF/BAT values, which errors were larger than threshold during curve fitting, was set to 0. The corrected CBF/BAT maps were first co-registered to the subject’s individual MPRAGE images and then normalized to the Montreal Neurological Institute (MNI) MNI152 T1 template. The normalized CBF/BAT maps in MNI coordinate system were averaged among all healthy subjects to create the normal CBF/BAT template. The CBF and BAT values of different brain areas were measured on the template using manually drawn ROI. A voxel-based analysis was used to analyze the age (younger group, 14 volunteers from 23 to 26 years old; older group, 14 volunteers from 30 to 58 years old) and gender effects (female group, 14 volunteers from 23 to 53 years old, average 32 years old; male group, 12 volunteers from 24 to 57 years old, average 31.5 years old). The CBF and BAT of the cerebellar infarction patients were then compared with the template. The CBF obtained with single-TI ASL (sTI-ASL) protocol (TI=2000ms) were also compared to mTI-ASL in both health and patient group.

RESULTS:

The mean CBF and BAT values of the different blood flow areas are shown in Figs. 2 and 3, respectively. The quantitative CBF values in different areas are consistent with the results obtained with PET1. The posterior circulation territory has a longer BAT than the hemicerebrum. No significant differences were found when comparing young vs. old groups and female vs. male groups. Fig. 4 shows a decrease in CBF and increase in BAT over the lesion area (shown on T2-FLAIR), compared to the template. On the opposite side of the lesion, CBF and BAT were still in the same range as in normal subjects. Compared to CBF measured by mTI-ASL, CBF obtained by using sTI-ASL shows significantly smaller values in both healthy subjects (p<0.001) and patient (p<0.001) group.

DISCUSSION:

This study indicates that CBF and BAT of different blood flow areas are not uniform, so the sTI-ASL protocols cannot accurately estimate CBF for the whole brain, especially for areas (such as the posterior circulation territory) with longer BAT values. In this right-handed population, the left hemisphere has higher CBF than the right hemisphere´╝îand BAT of the left hemisphere is shorter than of the right hemisphere. We did not find any differences between different age and gender groups, so one standard template might be fit for evaluating CBF and BAT changes in patients.

CONCLUSION:

The mTI-ASL method could quantify CBF and BAT values of the whole brain in one exam. It showed better performance than sTI-ASL, especially in areas characterized by longer BAT. The normal-subject CBF/BAT template could be used to aid the identification of perfusion deficits.

Acknowledgements

This work was funded in part by the Key R&D program in Shandong Province. The support of doctors (Fei Gao, Cuiyan Wang) and students (Junli Wang, Bao Wang) is gratefully acknowledged.

References

[1] P. van Gelderen et al. Pittfalls of MRI Measurement of White Matter Perfusion Based on Arterial Spin Labeling. Magnetic Resonance in Medicine 59:788–795 (2008).

Figures

Fig.1. The template of CBF and BAT of normal subjects.

Fig.2. The CBF value of different perfusion areas supplied from different arteries in left and right hemisphere. **p<0.01.

Fig.3. The BAT value of different perfusion areas supplied from different arteries in left and right hemisphere. *p<0.05.

Fig.4. TheT2-FLAIR, CBF and BAT of a 61-year-old female who had cerebellum infarction attack for 20 days.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
1447