3-D Microwave Tomography Using the Soft Prior Regularization Technique: Evaluation in Anatomically Realistic...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Electromagnetics

Background:

• Accurate breast imaging is crucial for early cancer detection.
• Microwave tomography offers potential for non-ionizing breast imaging.
• Integrating complementary imaging modalities can enhance diagnostic capabilities.

Purpose of the Study:

• To fuse magnetic resonance imaging (MRI) breast images with microwave tomography data.
• To improve the accuracy of microwave tomography image reconstruction using spatial information from MRI.
• To evaluate the performance of a soft prior regularization technique in 3D breast geometries.

Main Methods:

• A soft prior regularization technique was developed to incorporate spatial information from MRI into microwave tomography reconstruction.
• Numerical experiments were conducted on 3D breast models derived from MRI data with varying densities and simulated tumors.
• The technique's performance was assessed across diverse breast shapes, sizes, and tissue property distributions.

Main Results:

• The soft prior technique reduced root mean square error in permittivity and conductivity by over 87%.
• Contrast between simulated tumors and fibroglandular tissue improved significantly (59% in permittivity, 192% in conductivity).
• The algorithm demonstrated robustness across complex geometries and tissue property variations.

Conclusions:

• The soft prior algorithm is effective and robust for 3D breast imaging.
• Incorporating MRI spatial information enhances microwave tomography accuracy.
• This fusion technique shows promise for improved breast cancer detection and characterization.