Synopsis

Keywords: Hybrid & Novel Systems Technology, Hybrid & Novel Systems Technology, high permittivity materials, SNR enhancement

We show that high permittivity materials (HPMs) can improve functional magnetic resonance imaging (fMRI) at 1.5T, increasing the receive sensitivity of a commercial multi-channel head coil. Both numerical simulations and experiment showed ~25% improvement in the signal-to-noise ratio (SNR) of a commercial head coil in the areas of interest when HPM blocks were placed in proximity. This led to an increase in the detected motor cortex fMRI activation volume by an average of 56%, thus resulting in more accurate functional imaging at 1.5T.

Introduction

Reliable determination of regions with neuron activation during fMRI may be hindered due to the electric noise, often comparable in magnitude to the blood oxygenation level-dependent (BOLD) contrast. Practically, this requires a higher SNR^1,2. To further improve the SNR of local volume coils, one can use artificial structures, such as HPMs^3,4 or metasurfaces⁵. In particular, HPM pads have already proven themselves as a method to extend the field of view (FOV) and to improve the detection of BOLD activation in the cerebellum at 7T⁶. Here, we investigate numerically and experimentally with healthy volunteers the effect of HPM pads on receive (Rx) sensitivity of a commercial multi-channel head coil to improve the detection of brain activation regions in the motor cortex fMRI at 1.5T.

Methods

Electromagnetic (EM) simulations were performed in CST Microwave Studio 2021. To evaluate the transmit (Tx) efficiency and SNR, the following setups were considered: a whole-body birdcage coil (BC) and a receive-only multi-channel head coil with and without the HPM blocks in the presence of a homogeneous head phantom (Fig. 1A,B). The BC loaded with the phantom was matched and tuned to the working frequency at 63.64 MHz in the absence of the HPMs, the same was done with the head coil. Four HPM blocks⁴ based on BaTiO₃ with permittivity ε=4500, tan δ=0.1 and dimensions of 57×71×13 mm³ were introduced into the models in single-side position relative to the phantom (Fig. 1B). MATLAB R2021b was used for SNR evaluation with passive blocks. In addition, to estimate the effect of the HPM blocks in fMRI, we performed EM simulations using a multi-slice head phantom. SNR maps were calculated without and with HPMs presence to obtain SNR gain maps for further fMRI simulations. Phantom imaging was performed on a whole-body 1.5T scanner (Magnetom Espree, Siemens) using the body BC in Tx mode and Siemens Head Matrix TIM coil in Rx mode. B₁⁺ and SNR maps were obtained using a double-angle method⁷ with two gradient echo acquisitions and using data from one of the previous examinations divided by the standard deviation (SD) of noise in an MR image with no RF excitation, correspondingly. The homogeneity of the B₁⁺-field distribution was calculated as the ratio of the SD of the angle value to its mean value. To investigate the radiofrequency safety with HPMs, simulations with a voxelized human body model placed in BC were performed. It was evaluated by as |B₁⁺|_RMS divided by max SAR_av.10g, where |B₁⁺|_RMS - the root mean squared value in the ROI, max SAR_av.10g – maximum local SAR averaged over 10 grams of tissues. The B₁⁺-field and SAR distributions were normalized to 1W of the total accepted power. To investigate the impact of HPM blocks on the detection of neural activation in the motor cortex, we numerically simulated an fMRI study with a right-hand finger-tapping task using MATLAB-based STANCE toolbox⁷. fMRI experimental data were acquired for seven healthy volunteers with a right-hand finger-tapping task. To exclude the experiment order bias, the fMRI scans for three volunteers were performed first without blocks, then with blocks; for the remaining four volunteers, inverse order was used. Both numerical and experimental data were analyzed using group-level paired T-tests between acquisitions with and without HPM blocks in the MATLAB-based SPM toolbox⁸.

Results

Figure 1C,D shows numerically calculated |B₁⁺|_RMS-maps for a homogeneous phantom placed inside BC without and with HPMs, correspondingly. HPM blocks led to a 12% Tx efficiency gain, whereas the SNR gain map demonstrated 24% enhancement in the ROI (Fig. 1F) in comparison to the reference case. Figure 2C,D showed an experimentally slight improvement of the Tx efficiency, insignificantly disturbing the B₁⁺ homogeneity with HPM blocks in comparison to the reference case (91.2% versus 92.1%). The SNR gain map has a similar area of increase at the side of the phantom near the HPMs. Figure 3 demonstrates |B₁⁺|_RMS and SAR_av.10g maps for the human voxel model without and with HPMs. Similar to the phantom simulation’s results, one can observe an increase in the mean |B₁⁺|_RMS values with the presence of HPM blocks. The SAR efficiency with blocks was 1.09 uT/√W/kg compared to the reference case of 1.04 uT/√W/kg. The activation difference map between cases with blocks and cases without blocks for all volunteers, and the highest SNR simulations show statistically significant activation regions in the motor cortex for both numerical (Fig. 4, top two rows) and experimental (Fig. 4, bottom two rows) data. Figure 5 summarizes the results of fMRI experiments for all volunteers. The mean increase in the number of detected activation was 55% (from -60% to +215%).

Discussion and conclusions

It was demonstrated that HPM blocks can inexpensively improve the detection of BOLD activation in lower clinical field MRI. Both simulation and experiment results showed an improvement in the Rx sensitivity of the dedicated RF coil, providing significant gain in detecting the BOLD contrast. This method can increase the utility of existing commercial head coils at 1.5T, leading to a more accurate mapping of functional brain regions.

Acknowledgements

The part of the work related to the electromagnetic simulations and phantom studies was supported by a grant from the President of the Russian Federation for scientific school НШ-2359.2022.4. The part of the research devoted to the fMRI simulations and in vivo imaging was carried out with the support of the Ministry of Education and Science of the Russian Federation (075-15-2021-592).