Accelerating implant RF safety assessment using a low-rank inverse update method

Area of Science:

• Biomedical Engineering
• Computational Electromagnetics
• Medical Imaging

Background:

• Medical metallic implants (orthopaedic, pacemakers) restrict MRI access due to safety concerns.
• Radio frequency (RF) field-induced tissue heating is a primary risk during MRI.
• Current RF safety assessments for implants are computationally intensive due to scale differences.

Purpose of the Study:

• To develop a faster computational method for RF safety assessment of medical implants.
• To address the computational demands of RF safety analysis for detailed implant geometries.

Main Methods:

• Exploited implant geometry to develop a faster computational method.
• Utilized incident RF fields and a precomputed library matrix of implant edge responses.
• Employed a low-rank inverse update (Sherman-Woodbury-Morrison identity) for rapid EM response calculation.

Main Results:

• Achieved numerically equivalent electric and magnetic fields compared to full-wave simulations (max error 1.35%).
• Demonstrated significant speed improvements, 171 to 2478 times faster than GPU-accelerated simulations.
• Validated the method for two distinct implant geometries.

Conclusions:

• The presented method allows for rapid and efficient RF field evaluation near implants.
• This technique may enable customized MRI scanning conditions for patients with implants.
• Enhanced computational efficiency improves the feasibility of MRI for patients with metallic implants.