Introduction

Most 7T MRI sites have two head coils available for neuroimaging: a 1Tx head-coil for the single transmission (1Tx) system and an 8Tx head-coil for the parallel transmission (8Tx) system. In the circularly polarised (CP) mode (also known as TrueForm) the 8Tx coil should be identical to its 1Tx counterpart. However, no study has assessed this. Thus, we want to compare the performance of both systems for some commonly used sequences1 to determine any potential impact of switching routine 7T neuroimaging from 1Tx to 8Tx system in CP mode.

Methods

Six volunteers (2 female; age = 31.3±4.5 years; BMI=21.3±1.2 kg/m²) were scanned on a MAGNETOM Terra 7T MRI (Siemens Healthineers). Two scans were performed per participant: one on the 1Tx system with the 1Tx32Rx head coil (Nova Medical, USA) and another on the 8Tx system in CP mode with the 8Tx32Rx head coil (Nova Medical, USA). The CP mode’s per-channel complex weightings should give similar B₁⁺ to the birdcage 1Tx coil. Thus, both coils should give similar B₁⁺ spatio-temporal waveforms.
Scans were performed on the same day with a 15-minute break to allow switching between the systems. The time taken for the switch (inclusive of coil change) was also recorded.

Data acquisition: Six imaging sequences were acquired: dual-echo GRE for B₀-mapping, turbo-FLASH for B₁⁺-mapping, MP2RAGE, multi-echo GRE, GRE-EPI (5 subjects), and DTI with SE-EPI (3 subjects). The transmit voltage was determined automatically by the scanner. B₀-shimming was performed using the scanner’s auto-calibration routine and manual tuning of shim currents to minimise measured linewidths. All imaging scans had the same shim volume. Figure 1 shows the scanning parameters used.

Data analysis: Data analysis was conducted using FSL (v6.0.3, FMRIB, UK)² unless otherwise specified, and the workflow is shown in Figure 1. All ROI analyses were done in the native imaging space. A paired two-tailed Student’s t-test was used for statistical testing.

Results

On average, switching from 1Tx to 8Tx took 6:30 min and switching from 8Tx to 1Tx took 5:58 min.
Histograms of whole-brain B₀ and B₁⁺ values are presented in Figure 2. There were no significant differences in the mean and standard deviations (std) of B₀. There is a split peak in the 8Tx B₀-histogram for subject 2 due to a possible operator’s error when performing manual shimming. We saw significantly higher mean-B₁⁺ values with the 8Tx system in the caudate, insula, putamen, and temporal lobes. This improvement did not come with a penalty on the homogeneity: the 8Tx system showed significantly lower std-B₁⁺ in six out of nine ROIs.

Both coils showed similar MP2RAGE image quality in the cerebrum, with no significant differences in grey, white matter, and ventricular volumes (Figure 3). However, differences were seen in the brainstem and cervical spine. The spinal cord was only defined in 50% of the subjects with the 1Tx system, while clear boundaries between the spinal cord and spinal canal were seen in all volunteers with the 8Tx system. Figures 4A, B and C show the EPI tSNR maps with the mean and std-tSNR measurements. There were no significant changes in tSNR in any ROIs. Considering homogeneity in tSNR, the 8Tx system showed significantly higher std-tSNR in four basal-ganglia ROIs but significantly lower std-tSNR in the occipital-lobe ROIs. In terms of R₂*, the largest differences were observed in vessels and edges of the cortex (Figure 4D and E), but no significant differences in the mean and std-R₂* values were detected.

DTI images are shown in Figure 5. The b=0s/mm² images show a reduction of dropouts in the temporal lobes and cerebellum using the 8Tx system. The increased signal in those areas translates into more defined fitted white matter tracts.

Discussion

The two coils are often considered to be approximately equivalent (in CP mode). However, it is clear that the 8Tx coil actually provides better B₁⁺ efficiency in posterior brain regions. This translates into improved image quality for MP2RAGE and DTI without compromising other imaging sequences.
Switching to 8Tx mode and back took on average 12:28 min in total. With careful scheduling, it is feasible to run routine research protocols on the 8Tx system. Running protocols on the 8Tx system also opens up the possibility of including B1-shimmed/parallel transmission (pTx) protocols as add-on scans.
Although scanning on the 8Tx host does show some advantages, there are issues to consider before fully migrating protocols onto it. For example, changes in DTI signal intensity in the temporal lobes might affect a study if data acquired from the two different coils must be pooled. A future intra-coil and inter-coil reproducibility study will be needed to understand whether data can be pooled from both coils.

Conclusion

Overall, both systems give comparable image quality for routine neuroimaging (MP2RAGE structural, GRE-EPI, ME-GRE and DTI). The 8Tx coil gives more B1+ in the posterior brain regions which could benefit studies targeting those areas. It also allows pTx-enabled sequences to be run as an “add-on” to standard clinical protocols. Thus, we believe other sites could usefully investigate migrating their routine neuroimaging protocols to the 8Tx coil.

Acknowledgements

BD is funded by Gates Cambridge. CTR is funded by the Wellcome Trust and Royal Society (098436/Z/12/B). This research was supported by the NIHR Cambridge Biomedical Research Centre (BRC-1215-20014). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.