Diffusion PDE-based denoising technique for magnetic resonance electrical impedance tomography

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Electrical Engineering

Background:

• Magnetic Resonance Electrical Impedance Tomography (MREIT) shows promise for high-resolution conductivity imaging.
• A key challenge in MREIT is the need to limit injection current due to electrical safety regulations, resulting in low signal-to-noise ratio (SNR) data.
• Noisy MREIT data can degrade reconstructed conductivity images due to numerical differentiation in algorithms.

Purpose of the Study:

• To develop and evaluate a novel denoising technique for MREIT.
• To address the challenge of low SNR data caused by limited injection current.
• To improve the accuracy and preserve features in MREIT conductivity images.

Main Methods:

• Proposed a Partial Differential Equation (PDE)-based denoising technique.
• Applied the technique to mitigate noise in MREIT measured data.
• Evaluated the method's effectiveness in reducing random noise while preserving image features.

Main Results:

• The PDE-based denoising technique effectively reduces random noise in MREIT data.
• The method diminishes the degradation of reconstructed conductivity images caused by noise.
• Useful image features are preserved during the denoising process.

Conclusions:

• The proposed PDE-based denoising is a viable solution for improving MREIT image quality.
• This technique helps overcome limitations imposed by electrical safety regulations on injection current.
• Enhanced MREIT conductivity images are achievable, paving the way for potential clinical applications.