5014

Quantifying the Effects of Pulsed Saturation Transfer at 1.5T and 3T
Rachel W. Chan1, Sten Myrehaug2,3, Greg J. Stanisz1,4,5, Arjun Sahgal1,2,3, and Angus Z. Lau1,4

1Physical Sciences, Sunnybrook Research Institute, Toronto, ON, Canada, 2Radiation Oncology, Sunnybrook Health Sciences Centre, Toronto, ON, Canada, 3Department of Radiation Oncology, University of Toronto, Toronto, ON, Canada, 4Medical Biophysics, University of Toronto, Toronto, ON, Canada, 5Department of Neurosurgery and Pediatric Neurosurgery, Medical University, Lublin, Poland

Synopsis

The aim of this study was to investigate the feasibility of CEST imaging at 1.5T and to compare the results to 3T experiments. CEST and MT parameters were quantified in phantoms and in the healthy brain at both fields. A pulsed saturation scheme was used to overcome the single RF amplifier duty cycle limitations of the 1.5T clinical scanner. Parameters were estimated using a Bloch-McConnell two-pool simulation that incorporated the extended phase graph formalism. The new methods demonstrated promise for enabling broader application of CEST MRI for field strengths below 3T.

Introduction

The majority of chemical exchange saturation transfer (CEST)1 studies in the brain2-6 have been conducted at field strengths of 3T or higher, which provide increased spectral separation, reduced direct water saturation and higher SNR. However, the clinical relevance of saturation transfer techniques would benefit from the application of this methodology at lower field strengths. The aim of this study was to quantify the effects of CEST amide proton transfer (APT) in phantoms and in the healthy brain at both 1.5T and 3T. A pulsed saturation transfer scheme7-9 was used and recent technical advances were implemented to enable quantitative saturation transfer MRI on a typical 1.5T system.

Methods

MR protocol:
To overcome the RF amplifier limitations of a clinical 1.5T scanner, a novel pulsed saturation scheme (called pulse type “S”, for “short”), shown in Figure 1b, was used. At 3T, two saturation schemes, i) pulse type S and ii) a pulsed scheme with individual elements with longer durations (>200ms) (called pulse type “L”, for “long”, as in Mehrabian et al.2 and as shown in Figure 1a), were compared. MT/CEST sequence parameters are listed in Figure 2. A T1-weighted fast field echo (FFE) sequence was used for T1 mapping and a T2-weighted multi-echo sequence was used for T2 mapping. The WAter Shift And B1 (WASABI) sequence10 was used for B0 and B1 mapping.

Phantom experiments:
The phantom consisted of samples with varying NH4Cl concentrations (0M, 0.25M, 0.5M and 1.0M) to represent CEST contributions, mixed with 2% agar for the MT pool. The same phantom was imaged on the 1.5T (Philips Ingenia) and 3T (Philips Achieva) systems. On the 3T system, pulse types S and L were compared. MT and CEST parameters were compared across concentrations of NH4Cl. The CEST pool parameters were estimated for the tube with 1.0M NH4Cl for all three datasets.

In vivo experiments:
Informed consent was obtained and in vivo experiments were performed in the same healthy volunteer at 3T and at 1.5T in the brain. MT and CEST parameter maps were compared between the different field strengths.

Parameter Estimation:
Z-spectra were fitted to the Bloch-McConnell (BM) equations11 with extended phase graphs (EPG)12-13 incorporated in the model to keep track of magnetization in higher order echoes14. For MT quantification, the data were fitted using a two-pool model11 (i.e., liquid and semi-solid pools). The CEST effect was quantified by computing the difference between the CEST data and the quantitative MT-extrapolated curve15. The CEST asymmetry was also computed16.

Results and Discussion

Figure 3 shows example fits using a pulsed saturation scheme (pulse type S) at 1.5T. The model was well-fitted to the data points. In Figure 4, plots of the estimated MT and CEST parameters are shown from phantom experiments comparing two different pulse types (L and S), both implemented on the 3T system. Larger standard deviations in the MT parameters were seen using pulsed scheme S. For CEST, both the amide CEST effect and asymmetry increased with increasing NH4Cl concentration. For the CEST parameter fits, there was agreement in the relative CEST pool fraction, M0C, between 1.5T and 3T (using pulse type S), with M0C=4.1%(CI=3.9,4.2) at 1.5T and M0C=3.8%(CI=3.3,4.1) at 3T. The exchange rate between the water and CEST pools, RC, was smaller at 1.5T, with RC=36Hz(CI=33,38) at 1.5T and RC=53Hz(CI=44,62) at 3T.

Example maps of the healthy brain are shown in Figure 5 at 1.5T and 3T. Differentiation of white and gray matter was seen in the healthy brain at both field strengths, with improved white/gray matter contrast at 3T. There are notable differences in the CEST asymmetry (i.e. more negative asymmetry at 3T), and a larger nuclear Overhauser enhancement (NOE) contribution at 3T compared to 1.5T using the sets of B1 amplitudes that were tested.

Conclusion

Although only short pulse durations could be implemented on the 1.5T scanner due to RF amplifier limitations, the novel pulsed saturation scheme involving short 0.5ms pulses in the present study is a viable alternative method for saturation transfer and is suitable for systems that do not have dual transmit. The new methods demonstrated promise for enabling broader application of CEST MRI for field strengths below 3T and on systems with a single RF amplifier.

Acknowledgements

The authors acknowledge funding from NSERC, Terry Fox and CIHR.

References

1. Wolff SD, Balaban RS. NMR imaging of labile proton exchange. Journal of magnetic resonance. 1990;86:164-9.

2. Mehrabian H, Myrehaug S, Soliman H, Sahgal A, Stanisz GJ. Evaluation of Glioblastoma Response to Therapy With Chemical Exchange Saturation Transfer. Int J Radiat Oncol Biol Phys. 2018 Jul 1;101(3):713-723. doi: 10.1016/j.ijrobp.2018.03.057.

3. Mehrabian, H., Desmond, K. L., Soliman, H., Sahgal, A. & Stanisz, G. J. Differentiation between Radiation Necrosis and Tumor Progression Using Chemical Exchange Saturation Transfer. Clin. Cancer Res. clincanres.2265.2016 (2017). doi:10.1158/1078-0432.CCR-16-2265

4. Jones CK, Schlosser MJ, Van Zijl PC, Pomper MG, Golay X, Zhou J. Amide proton transfer imaging of human brain tumors at 3T. Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine. 2006 Sep;56(3):585-92.

5. Regnery S, Adeberg S, Dreher C, Oberhollenzer J, Meissner JE, Goerke S, Windschuh J, Deike-Hofmann K, Bickelhaupt S, Zaiss M, Radbruch A. Chemical exchange saturation transfer MRI serves as predictor of early progression in glioblastoma patients. Oncotarget. 2018 Jun 19;9(47):28772.

6. Jiang S, Eberhart CG, Lim M, Heo HY, Zhang Y, Blair L, Wen Z, Holdhoff M, Lin D, Huang P, Qin H. Identifying Recurrent Malignant Glioma after Treatment Using Amide Proton Transfer-Weighted MR Imaging: A Validation Study with Image-Guided Stereotactic Biopsy. Clinical Cancer Research. 2018 Jan 1:clincanres-1233.

7. Graham SJ, Henkelman RM. Understanding pulsed magnetization transfer. J Magn Reson Imaging 1997;7:903–912

8. Sled JG, Pike GB. Quantitative imaging of magnetization transfer exchange and relaxation properties in vivo using MRI. Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine. 2001 Nov;46(5):923-31.

9. Portnoy S, Stanisz GJ. Modeling pulsed magnetization transfer. Magnetic Resonance in Medicine: An Official Journal of the International Society for Magnetic Resonance in Medicine. 2007 Jul;58(1):144-55.

10. Schuenke, P., Windschuh, J., Roeloffs, V., Ladd, M. E., Bachert, P. & Zaiss, M. Simultaneous mapping of water shift and B1(WASABI)-Application to field-Inhomogeneity correction of CEST MRI data. Magn. Reson. Med. 77, 571–580 (2017).

11. Henkelman RM, Huang X, Xiang QS, Stanisz GJ, Swanson SD, Bronskill MJ. Quantitative interpretation of magnetization transfer. Magnetic resonance in medicine. 1993 Jun;29(6):759-66.

12. Weigel, M. Extended phase graphs: Dephasing, RF pulses, and echoes - pure and simple. J. Magn. Reson. Imaging 41, 266–295 (2015).

13. Hennig J. Echoes - how to generate, recognize, use or avoid them in MR‐imaging sequences. Part I: Fundamental and not so fundamental properties of spin echoes. Concepts in Magnetic Resonance Part A. 1991 Jul 1;3(3):125-43.

14. Malik, S. J., Teixeira, R. P. A. G. & Hajnal, J. V. Extended phase graph formalism for systems with magnetization transfer and exchange. Magn. Reson. Med. (2017). doi:10.1002/mrm.27040

15. Heo HY, Zhang Y, Jiang S, Lee DH, Zhou J. Quantitative assessment of amide proton transfer (APT) and nuclear overhauser enhancement (NOE) imaging with extrapolated semisolid magnetization transfer reference (EMR) signals: II. Comparison of three EMR models and application to human brain glioma at 3 Tesla. Magnetic resonance in medicine. 2016 Apr 1;75(4):1630-9.

16. Desmond KL, Stanisz GJ. Understanding quantitative pulsed CEST in the presence of MT. Magnetic resonance in medicine. 2012 Apr 1;67(4):979-90.

Figures

Figure 1 – Pulsed saturation modules for CEST: Two saturation pulse types for CEST are illustrated. a) Pulse type “L” (for “long”) consisted of four block pulses with relatively long durations (>200ms), separated by spoilers. b) The saturation module for pulse type “S” (for “short”) consisted of nine repetitions of Fermi pulses, with gradient spoiling in between the Fermi repetitions. Each Fermi pulse consisted of short 0.5ms RF blocks separated by 0.5ms gaps. Both pulse types were implemented on the 3T scanner. The 1.5T scanner allowed only pulse type S to be implemented.

Figure 2 – Saturation pulse sequence parameters for phantom experiments: MT and CEST sequences are shown for phantom experiments performed at 1.5T (with pulse type S) and 3T (with pulse types S and L). The B1 amplitudes represent the peak amplitudes, in μT. The reference frequency is represented by “ref”, followed by the multiplicative factor N. For example, “ref×N” indicates that the frequency of 100000Hz is repeated N times before subsequent saturation frequencies. The colon separates the start, step and end frequencies.

Figure 3 – Fitted MT and CEST curves at 1.5T: a) Phantom data (dots) and model fits (dashed lines) are shown for 2% agar with 1.0M of NH4Cl, acquired using pulse type S. Boxes 1 and 2 show the model fits (solid lines) simulated at closely spaced frequencies to illustrate the signal oscillations. In Box 1, both the model and data points are shown (1510-1690Hz). In Box 2, the model was evaluated from 0 to 2000Hz, every 5Hz. b) Extrapolated qMT curves (solid lines) and acquired CEST spectra (dots) are shown, with the differences plotted for each B1 power.

Figure 4 – Comparison of long and short pulsed saturation schemes at 3T: The MT and CEST parameters were compared between two pulse types, a) long (L) and b) short (S) for all ROIs with different concentrations of NH4Cl with 2% agar. The CEST effect and asymmetry were compared at the B1 amplitudes that generated the maximum values, with B1=1.0μT for pulse type L and B1,RMS=1.0μT for pulse type S.

Figure 5 – Saturation transfer maps at 3T and 1.5T in the healthy brain: Maps are shown of a similar slice in the healthy brain (of the same volunteer), scanned at a) 3T and b) at 1.5T. Saturation pulse type L was used at 3T and pulse type S was used at 1.5T. The CEST maps have been interpolated to B1=0.6μT for 3T and B1,RMS=1.5μT for 1.5T.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
5014