Introduction

Using an MRI-compatible research ultrasound (U/S) probe on a whole-body MRI system, simultaneous MR and U/S imaging has been demonstrated for the purpose of mitigating the effects of motion on the accuracy of image-guided therapy^1,2. MRI scanners equipped with high-performance gradients (HPG) combine high gradient amplitudes (G_max) and high maximum slew rates (SR_max) to enable high-resolution MRI with reduced susceptibility artifacts^3–5. For example, the MAGNUS gradient system can simultaneously deliver a G_max and SR_max of 300 mT/m and 750 T/m/s respectively³. However, multimodal MR and U/S imaging on an HPG MRI system remains unexplored. Here, we assess MR and U/S image quality, and the temperature rise in an MRI-compatible U/S probe at conventional gradient (CVG) and HPG configurations.

Materials and Methods

A multi-tissue U/S phantom (Model 040GSE, Zerdine^® Hydrogel) was used for all MR and U/S imaging.

Ultrasound Imaging
All U/S imaging was carried out on a 4D system (Vivid E95, GE Healthcare) equipped with a hands-free MR-compatible E4D research probe¹. 4D U/S images of the phantom were acquired in harmonic imaging mode (1.7/3.3 MHz) as shown in Figure 1A.

Temperature Measurement
The U/S probe surface temperature was measured during multimodality imaging using fiber optic temperature sensors placed on the surface of the probe (Figure 1B). A 10-minute wait time was allowed between each temperature measurement to permit convective cooling of the U/S probe. All temperature measurements were normalized to start at the ambient scanner room temperature (~21^oC).

Magnetic Resonance Imaging
All MRI HPG experiments were carried out on a MAGNUS gradient system (GE Healthcare, Waukesha, WI) at G_max=300 mT/m and SR_max=750 T/m/s. Imaging at CVG configuration was achieved by derating the MAGNUS system to operate at G_max and SR_max of 100 mT/m and 200 T/m/s respectively. RF signal transmit and reception were achieved using the integrated body coil and a 20-channel AIR anterior coil array (GE Healthcare) respectively (Figure 1C). To assess the B₀ inhomogeneity induced by the U/S probe, a 2D B₀-field map was obtained during U/S imaging using the following parameters: FOV=25 cm, TE/TR=3.4 ms/100 ms, voxel size=1.95x1.95x5 mm³, slices=40, average(s)=1. A control B₀ map without the U/S probe was also acquired with identical parameters. For U/S probe surface temperature measurement and image quality assessment, multi-phase fast spoiled gradient echo (FSPGR) images were obtained at CVG and HPG configurations for a duration of 5 mins with parameters: FOV=250 mm, voxel size=1.95x1.95x5 mm³, slices=10, TE/TR=0.9 ms/3.0 ms, bandwidth=±31.25 kHz, temporal resolution=4 s. EPI scans with identical parameters were obtained at bandwidth=±500 kHz using HPG (TE/TR=8.9 ms/32.0 ms, temporal resolution=0.32 s and echo spacing=0.3 ms) and CVG (TE/TR=14 ms/44.0 ms, temporal resolution=0.44 s/phase and echo spacing=0.6 ms) configurations.

Results and Discussion

Representative slices from B₀ maps and their corresponding magnitude images show regions of higher B₀ inhomogeneity around the location of the U/S probe (Figure 2A and 2B). The regions of higher inhomogeneity are not visible in the control images as the U/S probe was removed from the phantom. The depth of the B₀ inhomogeneity on the phantom was maximum (19.5 mm) at 28 mm from the gradient isocenter (Figure 2C). FSPGR and EPI MRI images obtained during temperature mapping on the surface of the U/S probe at CVG and HPG configurations are shown in Figure 3. There was no visible difference between the FSPGR images obtained at CVG and HPG configurations (Figure 3A). However, EPI distortion artifacts in the acquisitions at CVG configuration were more severe compared to HPG configuration due to the shorter ESP enabled by the latter (Figure 3B). There is also no significant difference in the U/S images obtained during multimodal imaging at CVG and HPG configurations (Figure 4). The temperature changes during the multimodal scans stayed below 28 ^oC except for the EPI scans carried out at HPG configuration (Figure 5). The U/S probe surface temperature reached ~38 ^oC at the sides of the probe. This increased temperature can be associated with higher gradient-induced eddy currents on metallic components in the U/S probe at higher SR_max. The eddy current heating at HPG configurations needs to be addressed in future work to fully utilize the higher SR_max of HPGs for EPI acquisitions during simultaneous MR-U/S imaging.

Conclusion

High-performance gradient systems can be utilized for rapid simultaneous MR-U/S FSPGR and EPI imaging with lesser distortion compared to conventional gradient configurations. However, there is a need for mitigation of eddy current induced heating in EPI at High-performance gradient configuration.

Acknowledgements

Research reported in this work was supported by the National Institute of Health under award numbers R01CA266879 and R01CA190298.