Introduction

Due to the barrier effect of the skull and cerebrospinal fluid, magnetic resonance elastography (MRE) faces challenges in generating usable shear waves in deep brain tissues compared to other tissues[1]. To obtain reliable whole-brain MRE data, appropriate scan parameters and image quality control are crucial and fundamental. Enhancing the propagation efficiency of shear waves in deep tissues, improving the sensitivity of motion-sensitive sequences to small shear wave displacements, optimizing post-processing, and inversion algorithms can enhance the quality of MRE images and increase the reliability of the results[2]. Gradient strength（GS） is one of the important factors affecting MRE images[3]. This study aims to investigate the impact of gradient strength on whole-brain MRE images, aiming to improve the quality of whole-brain MRE images and the credibility of the results.

Method

Five participants, including 2 males and 3 females, with an average age of 23±1 years, were enrolled in this study. All participants underwent MR imaging using a 3.0T scanner (Ingenia CX, Philips, the Netherlands) with 32-channel head coil. The whole scanning protocols were including structural T1, and 3D-MRE. The MRE excitation equipment was brought by Resoundant (Mayo Clinic, Rochester, MN, USA). The dedicated sequence and post-processing software were provided by Mayo Clinic and Philips. The gradient strength is set to 36mT/m or 70mT/m, other parameters of the 3D 3D-MRE images were as follows: motion encoding in the positive and negative x, y, and z directions; 8 phase offsets sampled over one period of the 60 Hz; 3 mm isotropic resolution; TR/TE = 4900/90 ms; FOV=240mmx240mm; 48 contiguous 3 mm thick axial slices. Figure 1 shows the specific post-treatment process for shear stiffness. The confidence map is an additional image obtained through post-processing in MRE, used to assess the reliability of data and analyze its region. A higher value typically indicates greater confidence in the shear stiffness results obtained through post-processing. Quantitative analysis of the confidence map is conducted using post-processing methods similar to those used for shear stiffness. Masks of Regions of Interest: Segmentation of gray matter and white matter was performed by using SPM12. The yield probability maps were binarized to obtain Global White Matter (GWM) and Cortical Gray Matter (CGM) mask, respectively. Regions of interest (ROIs) of cortical brain region were extracted and combined with Desikan–Killiany–Tourville atlas, including subcortical gray matter (CGM), frontal cortex (FC), parietal cortex (PC), temporal cortex (TC), and occipital cortex (OC). Subcortical Gary matter ROIs such as amygdala (AM), caudate (CA), hippocampus (HC), pallidum (PA), putamen (PU), thalamus (TH), and Accumbens (AC) were extracted from the HarvardOxford-sub -1mm atlas.

Result

Figure 2 illustrates the shear stiffness map, z-axis wave image, confidence map, and binary confidence map under different gradient strengths，reveals that at a gradient strength (GS) of 70mT/m, clear and continuous shear wave propagation of z-axis shear waves was observed in deep brain tissues, accompanied by a slight increase in shear stiffness values compared to GS of 36mT/m; The confidence map indicated significantly higher confidence values for the entire brain at 70mT/m compared to 36mT/m. After binarization of the confidence map, a larger analyzable area was observed at 70mT/m, allowing for more accurate and analyzable data. Quantitative analysis of the confidence map showed a significant improvement in confidence values for all regions of interest (p<0.05). Therefore, GS of 70mT/m yielded more accurate data. Quantitative analysis of the shear stiffness map also revealed higher parameter values, attributed to the influence of gradient field strength on the amplitude of phase information generated in magnetic resonance elastography, thus affecting the final inversion results.

Disscussion

The inclusion of motion encoding gradients (MEG) in specific MRI sequences is fundamental for recognizing and capturing tiny displacements generated by harmonic motion, forming the basis of MRE imaging[4]. Achieving effective, clear, and continuous wave propagation in deep brain tissues is essential for accurate whole-brain biomechanical information inversion calculations[5]. A comparison of different gradient strengths showed that higher gradient field strength enhances MEG's efficiency in recognizing and capturing small displacements, resulting in clearer and continuous harmonic propagation in wave images. Theoretically, increasing gradient field strength during phase unwrapping and conversion processes can reduce the influence of background phase and noise phase on the real harmonic motion phase. Consequently, this leads to more accurate biomechanical feature information during the inversion calculation process. Obtaining analyzable and trustworthy data for the entire brain is challenging due to the brain's unique geometric characteristics and biological structural composition. Therefore, opting for higher gradient field strength in whole-brain MRE is advantageous for obtaining more accurate and reliable data.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (81971581, 82272618) . Please address correspondence to Jian Yang, e-mail: yj1118@mail.xjtu.edu.cn and Xianjun Li, e-mail: xianj.li@mail.xjtu.edu.cn., Miaomiao Wang, e-mail: wangmm407@163.com.