Introduction

Moving to ultra-high fields (≥7T), the inhomogeneity of both RF (B₁) and static (B₀) magnetic fields increases, which further motivates to design a realistic head-shaped phantom. Such phantoms provide images similar to the human brain and serve as a reliable tool for developing and examining methods in MRI. In this study, a 3D-printed head-shaped phantom with brain mimicking metabolites and subcutaneous lipid layer examined for 7T MRI and MRSI. The phantom was designed to resemble the brain with respect to B₀ and B₁ distributions, T₁/T₂ relaxation times, metabolite content, and the subcutaneous lipid layer. The MRS and MRSI pulse sequences require a set of RF pulses, including water and lipid suppression, as well as a set of refocusing pulses, which are prone to both B₁ and B₀ inhomogeneity. In this study, we examined the use of such phantom for ¹H MRS and MRSI and for optimization in case of a lipid layer vicinity. In addition, EPI acquisition used for fMRI scans can reach high SAR when whole brain coverage is of interest. Here we demonstrate that the Fat Suppression pulse flip angle can be optimized to minimize the lipid artifact and reduce the SAR

Methods

The head-shaped container in this work was based on the Martinos Center’s “MGH Angel 001” ^1,2. The phantom was designed to include three sub-sections – mimicking brain, muscle and lipid tissues^3,4. The inner compartment was filled with a brain-mimicking mixture. The outer compartment was divided into two sections – the bottom one mimics muscle tissue, and the top one the lipid precranial layer. Gentle rolling during the agarose mixture filling allowed to generate thin layers on the outer shell surface that resemble skin/muscle and peanut oil was used to fill the left space⁵. The metabolite phantom included the main brain mimicking metabolites – see for details Ref. 4. Single voxel spectroscopy (SVS) on human and phantom was examined with LCModel fitting. In addition, SVS on phantom at two locations - central and close to the lipid layer (in the “visual cortex”) was examined with different spectral suppression parameters. EPSI) scan was used for MRSI acquisition with two sets of spectral width (SW) and spatial resolution. The Set #1 scan parameters were: TR/TE 2000/18 ms, FOV 300x300 mm², in-plane resolution 4.3x4.3 mm2 (70x70 pixels), slice thickness 20 mm, single slice, esp 0.52ms, SW 960 Hz, total scan duration 2:20 minutes. The Set #2 scan parameters: TR/TE 2000/18 ms, FOV 260x260 mm², in-plane resolution 8.7x4.0 mm² (30x64 pixels), slice thickness 20 mm, single slice, esp 0.33 ms, SW 1500 Hz, total scan duration 2:08 minutes. Finally, EPI imaging with and without Fat Suppression was examined for lipid artifacts. The Fat Suppression flip angle was varied to minimize the lipid artifact while reducing the SAR.

Results

The comparison of human and phantom B₀ and B₁ maps is shown in Figure 1. Images of the B₀ maximal deviation projection clearly demonstrated the main inhomogeneity areas near nasal, eye and ear regions, that were similar both in the phantom and in the human brain results. The human B₁ coefficient of variation was 37% and 28% for sagittal and axial scans, respectively, and that of the phantom was 23% and 18%, respectively. T₁ and T₂ relaxation times of the “brain” compartment were estimated in the central area as 1160±35 ms and 57±2 ms; those of the “lipid” compartment were estimated as 426±1 ms and 145±1 ms, respectively. Figure 2 shows representing sagittal and axial images of the phantom. Figure 3 shows SVS spectra and corresponding LCModel fit. The average standard deviation (SD) of the LCModel fitting (excluding GABA) was 4.8% for phantom and 3.6% for human. Figure 4 shows the EPSI results. Set #1 targeted high spatial resolution, which required reduction of the SW. The limited SW (~1000Hz) results in curved baseline in both spectrum edges due to the water peak (Fig.7c). Set #2 acquired lower spatial resolution with large enough SW for 7T ¹H spectra. The figure shows the water magnitude images and the NAA and Cr images as well as representative spectra in voxels moving from the center of the phantom to the edge. Figure 5 shows EPI with varying Fat Suppression flip angle. For a flip angle of 80° a relatively low lipid artifact can be achieved (7%) while reducing the SAR by 15% , compared to the default 110°.

Discussion

3D printed phantoms of high interest for 7T imaging and spectroscopic imaging, where B₁ inhomogeneity affects the quality of water and lipid suppression pulses, and B₀ inhomogeneity influences both water and lipid contamination. The B₀ and B₁ distribution measured in the phantom “brain” were in good agreement with human brain distribution, which is an important feature for the practical usage of the phantom. MRS, MRSI and MRI applications were examined. SVS scans on the phantom demonstrated a capability to optimize the spectral suppression parameters (depending on the application) for maximal NAA peak or minimal lipid artifact. The ability to exploit such a phantom to examine MRSI parameters was also demonstrated. Finally, reduction of SAR while minimizing the lipid artifact in EPI acquisition was shown.

Acknowledgements

We are grateful to Dr. Assaf Tal’s lab for assistance in the brain-mimicking metabolites preparation and LCModel fitting, and to Tamar Hayon and Efrat Biton from the Bacteriology Unit at the Dept. of LSCF, for technical assistance and media preparation. We thank Shimon Banouz and Slava Kofman from Laser Modeling for their assistance in the 3D printing. Dr. E. Furman-Haran holds the Calin and Elaine Rovinescu Research Fellow Chair for Brain Research.

References

[1] Guérin, B., (2016) Magn. Reson. Med 76(2), 540-554, [2] https://phantoms.martinos.org/MGH_Angel_001, [3] https://github.com/RitaSchmidt/3DRealisticHeadPhantomInMRI. [4] Jona et. al, NMR BioMed 2020. [5] Hines, C. (2009) J. Magn. Reson. Imag. 30(5), 1215-1222