MARIE 2.0: A Perturbation Matrix Based Patient-Specific MRI Field Simulator

Area of Science:

• Medical Imaging
• Computational Physics
• Biophysics

Background:

• High static field Magnetic Resonance (MR) scanners offer high-resolution imaging but generate complex, patient-specific electromagnetic fields.
• These fields can be potentially harmful, necessitating safety checks.
• Current parallel transmitter optimization relies on general guidelines due to slow patient-specific field computation.

Purpose of the Study:

• To develop a computational method for rapid patient-specific optimization of MR scanner parameters.
• To reduce the simulation time for in-tissue field patterns in MR imaging.
• To enable faster and more precise MR safety assessments.

Main Methods:

• Utilized a novel compressed-perturbation-matrix technique for voxel-based field simulation.
• Integrated fast low-resolution tissue mapping with accelerated field simulations.
• Implemented the method in the open-source MR field simulator MARIE 2.0.

Main Results:

• Reduced MR field simulation time from minutes to seconds.
• Demonstrated the technique's effectiveness across various head, coil, and shield configurations.
• Enabled near-elimination of computational cost for complex coil inclusions.

Conclusions:

• The compressed-perturbation-matrix technique significantly accelerates patient-specific MR field simulations.
• This advancement facilitates rapid, patient-specific MR safety checks and transmitter optimization.
• The method holds promise for improving the safety and efficiency of high static field MR imaging.