Comparison of Experimental and Computational Evaluation of Gradient-Induced Vibrations on Conductive Materials in MRI

To compare experimentally measured gradient-induced vibrations in MRI with computational results obtained from multiphysics finite element simulations.

Gradient-induced vibrations in a 3T scanner were quantified on commercially pure Titanium (CpTi) and polyetheretherketone (PEEK) plates following the ISO/TS 10974:2018 methodology. Sinusoidal and trapezoidal sequences were used. Vibrations were measured using Laser Doppler Vibrometry. The laser beam was redirected with a prism mounted on a custom measuring bench, allowing the measurement of vibrations perpendicular to the static magnetic field. RMS $dB/dt$ and the magnitude of the static magnetic field were measured and used as inputs for numerical simulations. The experimental results were compared with simulations performed using a weakly coupled multiphysics approach that involved the resolution of both electromagnetic and linear elasticity equations by the finite element method.

Both experimental and computational approaches gave access to the distribution of vibration magnitudes on the tested plates. The average relative difference between measured and simulated displacements was $\pm 11.5\%$ across all frequencies and all types of sequences tested. The transmitted vibrations of the scanner accounted for less than $1\%$ of the difference in displacements. The locations of maximal vibration magnitude were consistent between experiments and simulations.

Gradient-induced vibrations were evaluated both experimentally and computationally, with good agreement. Observed differences allowed for the evaluation of setup limitations and inherent uncertainties. Based on these results, the proposed computational approach can be used with a good level of confidence to evaluate and predict gradient-induced vibrations of implantable medical devices.