Sensitive measurements of weak current-induced magnetic fields in a 3D volume for in-vivo magnetic resonance current density imaging (MRCDI) of the human brain at high resolution

Results for the first subject (exemplary given) for the direct comparison of the final version of our improved MRCDI measurement method with our original method. (a) The improved method significantly reduces the TES-induced magnetic field Bz,c  noise floors (0 mA results) and enhances the image quality and resolution of the Bz,c  measurements (1 mA A-P current injection). It also successfully resolves spurious Bz,c  variations occurring near ventricles. The Bz,c  simulations and measurements show similar distributions, and are in the same range. (b) SNR and image quality changes are more clearly observed in the current flow reconstructions based on Bz,c  measurements and simulations, where noise floors are strongly reduced for our improved method. The reconstructed current flows from the simulations are similar to the measurements. (c) The scatter plots of the inferior slice results clearly demonstrate the relevance of the improved measurement method. The fitted line (green) approaches the identity line (orange) that is the ideal case. Our improved method also resolves the cluster (red) corresponding to the artifacts near ventricles.
 

Effective use of the transcranial electrical stimulation (TES) against neuropsychiatric disorders requires accurate mapping of the currents in the brain. Magnetic resonance current density imaging (MRCDI) combines MRI with low-intensity TES (1-2 mA) to map current flow in the brain. However, the utility of MRCDI is still hampered by low measurement sensitivity and image quality.
We recently introduced a multi-gradient-echo-based MRCDI approach that has the hitherto best documented efficiency. We now advanced our MRCDI approach in three directions and validated them by phantom and in-vivo experiments: First, we verified the importance of optimum spoiling for brain imaging. Second, we improved the sensitivity and spatial resolution by using acquisition weighting. Third, we added navigators as a quality control measure for tracking physiological noise. Combining these advancements, we tested our optimized MRCDI method by using 1 mA TES for two different injection profiles.
For a session duration of 4:20 min, the new MRCDI method was able to detect TES-induced magnetic fields at a sensitivity level of 84 pT, representing a twofold efficiency increase against our original method. Comparison between measurements and simulations based on personalized head models demonstrated a consistent increase in the coefficient of determination of ∆R2 =0.12 for the current-induced magnetic fields and ∆R2 =0.22 for the current flow reconstructions. Interestingly, some of the simulations still clearly deviated from the measurements despite of the strongly improved measurement quality. This highlights the utility of MRCDI to improve head models for TES simulations.
The achieved sensitivity improvement is an important step from proof-of-concept studies towards a broader application of MRCDI in clinical and basic neuroscience research.

1.
Göksu C, Scheffler K, Gregersen F, Eroğlu HH, Heule R, Siebner HR, Hanson LG, Thielscher A.
Sensitivity and resolution improvement for in vivo magnetic resonance current-density imaging of the human brain.
Magn. Reson. Med. 2021;86:3131–3146. doi: 10.1002/mrm.28944.
 
Go to Editor View