Electrical properties based B 1 + prediction for electrical properties tomography reconstruction evaluation.

Area of Science:

• Medical Imaging
• Biophysics
• Computational Science

Background:

• MR electrical properties tomography (EPT) reconstructs conductivity and permittivity from MR measurements.
• Reconstructed EPs exhibit significant in vivo variability, limiting clinical trust.
• A novel physics-based estimation model is proposed to address this variability.

Purpose of the Study:

• To develop and validate a method for evaluating reconstructed electrical properties (EPs) in MR EPT.
• To enhance confidence in in vivo EPT results by assessing the accuracy of reconstructed EPs.
• To utilize a physics-based magnetic field ($$) estimation model for EP evaluation.

Main Methods:

• A finite difference-based recurrent relation estimates the $$ field from given EPs and boundary conditions.
• The method compares the estimated $$ field with the measured $$ field for EP evaluation.
• Validation was performed using simulations, phantom MRI data, and in vivo MRI data.

Main Results:

• Simulations demonstrated accurate $$ field estimation (90s for brain, 1mm³ resolution) with correct EPs.
• Discrepancies between estimated and measured $$ fields indicated errors in input EPs.
• MRI experiments confirmed the method's practical applicability in phantoms and in vivo.

Conclusions:

• The proposed method accurately estimates $$ fields from EPs.
• This approach provides a robust evaluation of EPT reconstructions.
• It increases confidence in the in vivo reconstructed EPs for clinical applications.