Reproducibility of Cerebral Blood Flow Measurements Across MR Systems: A Matter of Magnet Geometry
Bogdan G Mitrea1, Ralf B Loeffler1, Ruitian Song1, and Claudia M Hillenbrand1

1Diagnostic Imaging, St Jude Children's Research Hospital, Memphis, TN, United States

Synopsis

For longitudinal investigations reproducibility and accuracy of ASL measurements are of essence. The purpose of this study is to investigate how magnet design and labeling position within the magnet impact CBF quantification. Our results indicate that CBF values are not always reproducible in our ultra-short wide bore scanner. Great variability may be introduced by the actual position of the labeling slice with respect to the magnet isocenter. The exact cause of this difference requires further investigation. However, positioning the labeling slice in isocenter provided a simple solution to overcome this issue and to measure reproducible CBF values.

Purpose

Cerebral blood flow (CBF) is an important measure of brain function, often applied to diagnose and evaluate cerebral disorders1. Arterial spin labeling (ASL) is widely used for serial, noninvasive quantification of CBF to assess treatment response or disease progression2. For such longitudinal investigations reproducibility and accuracy of ASL measurements are of essence. Past ASL studies have shown reproducibility of CBF estimates3–5, but did not investigate how magnet design and labeling position with respect to the isocenter impact quantification. The purpose of this study is to investigate these potential issues.

Methods

Fifteen healthy volunteers enrolled in an IRB approved protocol were scanned back-to-back on two 3T systems, a standard bore Siemens Trio (M1), and an ultra-short wide bore Skyra (M2). Scanner software versions were B17 and D13, and respective product Q2TIPS6 ASL sequences were applied with the following parameters: TR=2280ms, TE=23ms, TI=1400ms, TI1=700ms, FOV=210×210mm2, matrix 64×64, 11 slices, thickness 5mm, and 1mm gap. The labeling slice was 100mm thick and placed 20mm inferior to the imaging block. Sequence code and parameters were reviewed by an experienced programmer and the timing parameters were found to be identical. ASL scans for each magnet were performed at three specific table positions: ISO, where the center of the imaging block was aligned with the magnet’s isocenter, ISO-50 and ISO-100, where the center of the imaging block was placed 50mm and 100mm superior from isocenter. In ISO-100 position, the center of the labeling slice coincided with the magnet’s isocenter. A 3D T1-weighted full brain volume was acquired and used for segmentation and registration using the FSL toolbox7. CBF was calculated8 using Matlab, assuming identical label efficiency independent of magnet type and table position. Two-way repeated measurement ANOVA was performed in SigmaPlot to test for difference in CBF quantification between MR systems and table position. To test the homogeneity of the main magnetic field, a B0 field map was obtained from a series of 3D multi-gradient echo phase images 9,10 (FOV=256×256mm2, matrix size =128×128, slice thickness 5mm, 32 slices, TR=50ms, 3 TEs of 6, 11, 16ms and FA=25°).

Results

Figure 1 shows representative gray matter (GM) CBF maps of the same volunteer at the three table positions acquired on both magnets. On M1, CBF maps appear approximately equivalent and consistently show the same results independent of the table position. M2 maps, however, demonstrate decreased CBF values when the labeling block moves away from isocenter. Only at ISO-100, where the labeling block is aligned with the isocenter, were CBF values comparable between M1 and M2. The results of the two-way repeated measure ANOVA (Figure 2) emphasize a difference in CBF quantification between magnets dependent on table position (p<0.01). The individual magnet analysis found that on M2, CBF values measured in ISO-100 are significantly different from those in ISO-50 and ISO (p<0.001). In contrast, CBF values from M1 were reproducible independent of table position. Further, we found a significant difference between M1 and M2 for CBF values measured at ISO and ISO-50 (p<0.01); however, no difference was observed for CBF values obtained at ISO-100 (p=0.349).

Discussion

Our results indicate that CBF values are not always reproducible. Great variability may be introduced by the actual position of the labeling slice with respect to the isocenter in our ultra-short wide bore M2 scanner. As shown in Figure 3, B0 field maps acquired at ISO and ISO-100 on M1 are virtually identical while the ones acquired on M2 show visible differences. This could indicate a less spatially homogenous field which decreases the labeling efficiency and results in a lower CBF. ASL images are typically acquired in ISO position which, in this study, produced the largest CBF bias on the short bore magnet (20% lower than M1). Moving the tagging slice into isocenter (ISO-100), where the field is most homogenous, resulted in the highest CBF values that are consistent between M1 and M2. Our results indicate that the labeling position in ASL studies should be fixed at isocenter to assure comparable CBF values for cohort studies or longitudinal exams that involve short bore magnets.

Conclusion

We noticed systematic differences in CBF values between a long and short bore magnet. The exact cause of this difference requires further investigation. However, positioning the labeling slice in isocenter provided a simple solution to overcome this issue and to measure reproducible CBF values on both magnets.

Acknowledgements

References

1. Alsop, D. C. et al. Recommended implementation of arterial spin-labeled perfusion MRI for clinical applications: A consensus of the ISMRM perfusion study group and the European consortium for ASL in dementia. Magn. Reson. Med. n/a–n/a (2013). doi:10.1002/mrm.25197

2. Chao, L. L. et al. ASL perfusion MRI predicts cognitive decline and conversion from MCI to dementia. Alzheimer Dis. Assoc. Disord. 24, 19–27 (2010).

3. Petersen, E. T., Mouridsen, K. & Golay, X. The QUASAR reproducibility study, Part II: Results from a multi-center Arterial Spin Labeling test–retest study. NeuroImage 49, 104–113 (2010).

4. Gevers, S. et al. Intra- and multicenter reproducibility of pulsed, continuous and pseudo-continuous arterial spin labeling methods for measuring cerebral perfusion. J. Cereb. Blood Flow Metab. 31, 1706–1715 (2011).

5. Mutsaerts, H. J. M. M. et al. Inter-Vendor Reproducibility of Pseudo-Continuous Arterial Spin Labeling at 3 Tesla. PLoS ONE 9, e104108 (2014).

6. Luh, W.-M., Wong, E. C., Bandettini, P. A. & Hyde, J. S. QUIPSS II with thin-slice TI1 periodic saturation: A method for improving accuracy of quantitative perfusion imaging using pulsed arterial spin labeling. Magn. Reson. Med. 41, 1246–1254 (1999).

7. Jenkinson, M., Beckmann, C. F., Behrens, T. E. J., Woolrich, M. W. & Smith, S. M. FSL. NeuroImage 62, 782–790 (2012).

8. Wong, E. C., Buxton, R. B. & Frank, L. R. A theoretical and experimental comparison of continuous and pulsed arterial spin labeling techniques for quantitative perfusion imaging. Magn. Reson. Med. 40, 348–355 (1998).

9. Webb, P. & Macovski, A. Rapid, fully automatic, arbitrary-volume in vivo shimming. Magn. Reson. Med. 20, 113–122 (1991).

10. Schneider, E. & Glover, G. Rapid in vivo proton shimming. Magn. Reson. Med. 18, 335–347 (1991).

Figures

Representative example of GM CBF map from two imaging slices for M1 and M2 at the three imaging positions. CBF is underestimated on M2 for ISO and ISO-50.

Position dependent differences in CBF quantification between the magnets. Asterisk indicates statistical significance (p<0.01) mean CBFs

Comparison of B0 field maps (center slice) between M1 and M2 at ISO and ISO-100 positions.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4087