1973

The comparison of different strategies for transmit field inhomogeneity correction of Amide-CEST and NOE effects at 7T

Vitaliy Khlebnikov¹, Johannes Windschuh², Jeroen CW Siero^1,3, Moritz Zaiss⁴, Peter R Luijten¹, Dennis WJ Klomp¹, and Hans Hoogduin¹

¹Department of Radology, University Medical Center Utrecht, Utrecht, Netherlands, ²Division of Medical Physics in Radiology, Deutsches Krebsforschungszentrum (DKFZ) [German Cancer Research Center], Heidelberg, Germany, ³Spinoza Center for Neuroimaging, Amsterdam, Netherlands, ⁴Scheffler, Max Planck Institute for Biological Cybernetics, Tübingen, Germany

Synopsis

We compared three methods for B1 correction of relaxation-compensated Amide-CEST and Nuclear Overhauser Enhancement (NOE) effects at 7T: (1) a linear model; (2) the eight-point interpolation method; and (3) Bloch-McConnell equations (BE) correction algorithm. In the low B1 regime of 0.10 - 0.50 µT, a simple linear model is sufficient to mitigate B1 inhomogeneity of Amide-CEST and NOE effects at 7T.

Target audience.

Those interested in the application of Chemical Exchange Saturation Transfer (CEST).

Introduction.

B₁ inhomogeneity correction is vital for high-field CEST-MRI applications. In this work, we compared three methods for B₁ correction of relaxation-compensated Amide-CEST and Nuclear Overhauser Enhancement (NOE) effects at 7T: (1) a linear model; (2) the eight-point interpolation method [1]; and (3) Bloch-McConnell equations (BE) correction algorithm.

Methods. Data acquisition.

The experiments were performed on a 7 T MR whole-body system (Siemens, Erlangen, Germany). All data was acquired from the same consented volunteer. Z-spectra were acquired at eight B₁ levels (average power [2] 0.14, 0.29, 0.43, 0.50, 0.58, 0.65, 0.72, and 0.80 µT) and sampled at 66 frequency offsets using a train of 120 (15ms, duty-cycle 60%) GR-spoiled Gaussian-shaped pulses for saturation. The readout was a single-shot 2D gradient-echo sequence (GRE) with GRAPPA factor 2, TR/TE/FA=7.4ms/3.6ms/10°, matrix 128x128, slice thickness 5mm. Total scan time was 4min07s [3]. WASSR method was used for B₀ correction [4]. A relative B₁ map was generated using a single-shot GRE sequence. White matter (WM) and grey matter (GM) masks were generated in FSL (FMRIB v6.0, UK). The apparent exchange dependent relaxation (AREX or relaxation compensated MTR_Rex) [5,6], was not calculated, since the B₁ dependency remains the same for MTR_Rex and AREX.

Methods. Data processing.

Z-spectra were fitted using six-pool (water, Amide-CEST, NOE, magnetization transfer, Amine-CEST [1] and NOE* [7] BE at a single B₁ level (the only fixed parameter in the fit and defined as B_1,nominal•B_1,relative) at any given time. Amide-CEST effect size was quantified using the inverse metrics to suppress the effects of direct water saturation and traditional magnetization transfer [5,6]:

MTR_Rex,Amide= 1/Lab - 1/Ref (1)

where Lab = M_z(Δω, M_b = 1)/M₀ and Ref = M_z(Δω, M_b = 0)/M0 are the signals in the Z-spectrum at Δω = 3.5 ppm, M₀ is the equilibrium magnetization (Δω = 500 ppm) and M_b is the amplitude (M_b = 0 and M_b = 1 for the system without and with Amide-CEST pool, respectively). A similar equation applies to the NOE pool (MTR_Rex,NOE) at Δω =-3.5 ppm. Both the linear and BE B₁ correction algorithms are explained in Fig. 1. The eight-point contrast-based interpolation B₁ correction (the golden standard used as a reference) was implemented by interpolating between all contrast maps (MTR_Rex,Amide or MTR_Rex,NOE) at all B₁ levels to a B₁ of 0.43 µT [1].

Results and Discussion.

All three B₁ correction algorithms were compared at a nominal B₁ level of 0.43 µT, the actual B₁ values (60 - 120% inhomogeneity) of which are in the linear regime (Fig. 2a). The linear relationship of both MTR_Rex,Amide and MTR_Rex,NOE metrics with B₁ in the low B₁ regime (0.1 – 0.5 µT) opens a possibility of a simple linear B₁ correction. In Fig. 3, BE B₁ correction was applied to the in vivo data. The CEST spectra were well fitted with BE (Fig. 3b). The overlap of BE B₁ corrected spectra (Fig. 3c) suggests that BE may alleviate B₁ inhomogeneity.The linear B₁ correction improved image quality and the B₁ corrected contrast (both MTR_Rex,Amide and MTR_Rex,NOE) resembles that of the interpolation B₁ correction method (the reference method), whereas BE B₁ correction resulted in over- and under correction at low and high B1, respectively (Fig. 4). In addition, the linear B₁ correction method virtually nullified contrast correlation with B₁, whereas BE B₁ correction clearly over-compensated contrast at low B₁ values (Fig. 5). The results of this work are also applicable to other methods used for the quantification of Amide and NOE CEST effects such as the three-point method [8] (data now shown) and multiple-Lorentzian fitting approach [1] (data not shown).

Conclusions.

In the low B₁ regime of 0.10 - 0.50 µT, a simple linear model is sufficient to mitigate B₁ inhomogeneity of Amide-CEST and NOE effects at 7T.

Acknowledgements

The financial support of the European Commission under FP7 Marie Curie Actions (FP7-PEOPLE-2012-ITN-316716) is gratefully acknowledged.

References

[1] Windschuh J, Zaiss M, Meissner JEE, Paech D, Radbruch A, Ladd ME, Bachert P. Correction of B1-inhomogeneities for relaxation-compensated CEST imaging at 7T. NMR in biomedicine. 2015;28(5):529–537.

[2] Zu Z, Li K, Janve VA, Does MD, Gochberg DF. Optimizing pulsed-chemical exchange saturation transfer imaging sequences. Magn Reson Med. 2011;66(4):1100–1108.

[3] Schmitt B, Zaiß M, Zhou J, Bachert P. Optimization of pulse train presaturation for CEST imaging in clinical scanners. Magn. Reson. Med. 2011; 65: 1620–1629.

[4] Kim M, Gillen J, Landman BA, Zhou J, van Zijl PC. Water saturation shift referencing (WASSR) for chemical exchange saturation transfer (CEST) experiments. Magn Reson Med. 2009;61(6):1441–1450.

[5] Zaiss M, Xu J, Goerke S, Khan IS, Singer RJ, Gore JC, Gochberg DF, Bachert P. Inverse Z-spectrum analysis for spillover-, MT-, and T1-corrected steady-state pulsed CEST-MRI – application to pH-weighted MRI of acute stroke. NMR Biomed. 2014;27(3):240–252.

[6] Zaiss M, Windschuh J, Paech D et al. Relaxation-compensated CEST-MRI of the human brain at 7T: Unbiased insight into NOE and amide signal changes in human glioblastoma. NeuroImage. 2015;112:180–188.

[7] Zhang XY, Xie J, Xu J, Li H, Gore JC, Zu Z. Assessment of membrane fluidity using nuclear overhauser enhancement mediated magnetization transfer (NOE-mediated MT). ISMRM 2015, Toronto, Canada. 1747.

[8] Jin T, Wang P, Zong X, Kim SG. MR imaging of the amide-proton transfer effect and the pH-insensitive nuclear overhauser effect at 9.4 T. Magn Reson Med. 2013;69(3):760–770.

Figures

Fig. 1. A flowchart explaining the implementation of a linear model and BE B₁ correction algorithms.

Fig. 2. The experimentally derived plots of MTR_Rex,Amide and MTR_Rex,NOE versus actual B₁ values in WM. The relative B₁ map was segmented into the regions between 50% and 150% in the steps of 1% and the corresponding MTR_Rex,Amide and MTR_Rex,NOE was calculated for all datasets. The linear regression analysis (straight black lines) was done in the B₁ range 0.1 - 0.5 µT. The Pearson's correlation coefficient (R) and the corresponding p-value are provided. ** represents statistical significance at the level p<<0.005.

Fig. 3. (a) A CEST image of the healthy human brain with the cross marking the origin of the CEST spectra shown in (b) and (c). (b) The in vivo CEST spectra (colored markers) at various B₁ levels and their corresponding six-pool BE fits (colored solid lines). (c) Same as (b) for the colored markers, but the colored solid lines represent BE corrected spectra recalculated at a B₁ of 0.43 µT (assumed to be nominal B₁ level) using the corresponding fitting parameters from (b).

Fig. 4. (a) A relative B₁ map. (b) The comparison of the experimentally derived uncorrected, the linear model B₁ corrected, the BE B₁-corrected and the interpolation based B₁ corrected contrast for MTR_Rex,Amide (top row) and MTR_Rex,NOE (bottom tow), respectively.

Fig. 5. The voxel-wise correlation of the image contrast (Fig. 4b) with the relative B₁ map (Fig. 4a). (a) and (b) Originate from WM and GM, respectively, for MTR_Rex,Amide. (c) and (d) Originate from WM and GM, respectively, for MTR_Rex,NOE. The linear colored lines represent the linear regression. The Pearson's correlation coefficient (R) is shown in each subfigure.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)

1973