Computational methods for the estimation of ideal current patterns in realistic human models.

Area of Science:

• Medical Imaging
• Electromagnetism
• Biophysics

Background:

• Optimizing signal-to-noise ratio (SNR) is crucial for high-quality magnetic resonance imaging (MRI).
• Realistic modeling of heterogeneous human tissues presents challenges for radiofrequency (RF) coil design.
• Current analytical and numerical methods for RF coil optimization have limitations in complex scenarios.

Purpose of the Study:

• To introduce a novel method for estimating ideal current patterns (ICPs) that maximize SNR in MRI.
• To apply this method to realistic heterogeneous human tissue models.
• To guide the design of application-specific RF coil arrays for enhanced MRI performance.

Main Methods:

• Calculated ICPs for surfaces mimicking typical RF coil formers.
• Constructed numerical electromagnetic (EM) models to represent RF current sources and fields.
• Solved volume integral equations to compute EM fields within tissue models and determined optimal SNR and ICPs.

Main Results:

• ICP shapes varied from distributed loops to figure-eight patterns depending on voxel location within the tissue model.
• An RF coil array inspired by ICPs achieved a significant fraction of optimal SNR at 3T, outperforming a standard single loop.
• The performance advantage of ICP-guided designs diminished at 7T, indicating potential limitations of loop designs at ultra-high fields.

Conclusions:

• The developed method successfully calculates ICPs for human tissue models.
• ICPs provide valuable insights for designing application-specific RF coil arrays.
• This approach holds potential for improving MRI performance across different field strengths.