Frequency correction based on interleaved water acquisition improves spectral quality in MM-suppressed GABA measurements in vivo
Nicolaas AJ Puts1,2, Kimberly L Chan1, Ashley D Harris1,2,3,4, Peter B Barker1,2, and Richard AE Edden1,2

1Radiology and Radiological Science, Johns Hopkins University, Baltimore, MD, United States, 2F.M. Kirby Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States, 3Alberta Children's Hospital Research Institute, University of Calgary, Calgary, AB, Canada, 4Radiology, University of Calgary, Calgary, AB, Canada

Synopsis

MM-suppressed GABA measurements use symmetric editing of both MM and GABA signals. Frequency drift, either by gradient induced heating/cooling, or motion, significantly affects the editing efficiency of GABA and MM. To stabilize the center frequency, we interleaved the unsuppressed water acquisition throughout the scan and used it to correct the frequency, in eight healthy participants, and compared this to a condition without frequency correction. Frequency correction improves spectral quality of MM-suppressed GABA editing in vivo.

Background

Typical GABA-edited 1H magnetic resonance spectroscopy suffers from as much as 50%1 macromolecular (MM) contamination of the GABA signal at 3 ppm, as the frequency selective editing pulse for GABA (at 1.9 ppm) also partially inverts MM signals (at 1.7 pm). Symmetrical editing, by applying the editing pulses around the 1.7 MM peak (at 1.9 and 1.5 ppm) allows for the suppression of unwanted MM signal as the MM signal is inverted to an equal degree in both the ON and OFF experiments2. However, field instability may occur during the course of an experiment due to patient movement and/or gradient-induced heating/cooling3. For MM-suppressed experiments specifically, field changes can substantially change the editing efficiency of the editing target molecule and result in imperfect symmetric suppression of nulled contaminants, which cannot be addressed by retrospective frequency-and-phase correction before signal averaging3-5. Frequency correction during scanning (in addition to post-hoc frequency correction) is therefore required. This abstract describes the implementation of a modified prospective frequency scheme which adds no additional scan time, and demonstrates its use in vivo under both low and high drift rate conditions.

Methods

MM-suppressed GABA MRS data were acquired on a Philips 3T Achieva scanner, in eight healthy participants (5F) with the following parameters; TE/TR 80/2000 ms; 20 ms editing pulses at 1.9 (ON) and 1.5 (OFF) ppm, 320 transients, 32 channel head coil. Measurements were made from (3.5cm)3 voxel placed in the midline parietal lobe. Two acquisitions were made for each participant, one with frequency correction on (FC-on) and one with frequency correction off (FC-off). Frequency correction was performed based on a periodically acquired water signal (without water suppression) from the localized volume: rather than acquiring 16 averages (one full phase cycle) of water reference data at the start of the acquisition (Fig. 1B), as is typically done for eddy-current correction6 and quantification.Instead these reference transients were interleaved throughout the acquisition (Fig 1A). Data were acquired under either high or low field drift conditions (high drift being induced in a single case by a preceding 1 hour EPI/DTI acquisition known to cause gradient heating, lower drift induced by a preceding 2 min DTI acquisition). For one participant, only FC-on was acquired. All data were processed using Gannet7, including a further post-hoc frequency correction5 to minimize subtraction artifacts. GABA levels were quantified relative to the unsuppressed water signal from the same volume.

Results

Comparison of in vivo data with- and without prospective frequency correction (FC-on and FC-off), show considerably more stable water frequencies with correction on (Fig 2AB). Fig 3C shows more variability in the water frequency drift in the FC-off condition (2.01 ± 0.43 Hz; without outlier 1.07 ± 0.52) than in the FC-on (0.81 ± 0.43 Hz; without the scan coupled to outlier: 0.61 ± 0.32 Hz). As shown in Figures 2A and 2B, there exists considerable drift after an hour of EPI/DTI imaging without FC (46 Hz), but is limited (and corrected) with frequency correction to a maximum of 6.4 Hz. Substantial drift leads to altered editing of both the GABA and MM peak (editing pulses are no longer on-resonance) leading to an apparently negative ‘GABA’ peak in this case (Figure 3B). While in the case of mild drift there are no significant differences in average GABA levels, GABA levels in the FC-off condition are more variable (standard deviation FC-off 0.23 and FC-on 0.16). The GABA levels are lowest for the scans after 2 min of DTI, and negative (not shown for FC-off) after 2hr of DTI/EPI (Figure 3D).

Discussion

Symmetric suppression is an elegant concept that allows an unwanted co-edited signal to be removed, augmenting the selectivity of the editing pulses themselves. This symmetry is broken by changes in the scanner frequencies that arise from subject movement or temperature-related scanner drift, leading to artifactual increases or decreases in the edited signal. GABA values tend to lower with increasing drift, and interpretation of the GABA peak when inappropriately edited, is problematic. By interleaving the water reference scans throughout the acquisition, the water frequency can be corrected prospectively, reducing the effect of drift of data quality (a maximum drift of 6.4 Hz corresponds to 0.04 ppm and it's well known editing efficiency drops off quickly off-resonance), and effect of drift of the efficiency of the editing pulses, considerably improving data quality.

Conclusion

Drift has a strong and immediate impact on MM-suppressed GABA MRS measurements. Given the strong reliance on applied editing pulses on resonance in symmetric MM-suppression, we do not recommend applying MM-suppression without the use of prospective frequency correction during scanning.

Acknowledgements

NIH grants: R01 EB016089, P41EB015909, R21 NS077300

References

1. Mullins PG, McGonigle DJ, O'Gorman RL, et al. Current practice in the use of MEGA-PRESS spectroscopy for the detection of GABA. Neuroimage 2014;86:43-52.

2. Edden RA, Puts NA, Barker PB. Macromolecule-suppressed GABA-edited magnetic resonance spectroscopy at 3T. Magn Reson Med 2012;68:657-661.

3. Harris AD, Glaubitz B, Near J, et al. Impact of frequency drift on gamma-aminobutyric acid-edited MR spectroscopy. Magn Reson Med 2014;72:941-948.

4. Evans CJ, Puts NA, Robson SE, et al. Subtraction artifacts and frequency (mis-)alignment in J-difference GABA editing. J Magn Reson Imaging 2013;38:970-975.

5. Near J, Edden R, Evans CJ, Paquin R, Harris A, Jezzard P. Frequency and phase drift correction of magnetic resonance spectroscopy data by spectral registration in the time domain. Magn Reson Med 2014.

6. Klose U. In vivo proton spectroscopy in presence of eddy currents. Magn Reson Med : 1990;14:26-30.

7. Lange T, Zaitsev M, Buechert M. Correction of frequency drifts induced by gradient heating in 1H spectra using interleaved reference spectroscopy. JMRI; 2011;33:748-754.

Figures

Figure 1. Water acquisitions are interleaved throughout the scan, and the center frequency is updated at each water frequency. B. During a typical scan, all water scans (a full phase cycle) are acquired at the start of the acquisition.

Figure 2. Water traces during scans with the prospective frequency correction on (A) and off (B). Center frequency is updated/corrected after each water acquisition in the FC-on condition. Substantially more drift can be seen when the correction is not used (B), especially in the case of substantial gradient use (AB).

Figure 3. Spectra for all participants. A. Good quality data were acquired for all participants in the prospective FC-on. B. More variability in spectral quality in FC-off. Drift can be so substantial that editing pulses are off-resonance and lead to 'negative-GABA signal' (B). FC-off was not acquired for one participant.

Figure 4. A. More variability in water frequency in the FC-off condition. Frequency standard deviation is also greater in the FC-off condition. B. GABA values are lower and more variable in the FC-off condition.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2385