Image reconstruction for positron emission tomography using fuzzy nonlinear anisotropic diffusion penalty

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Science

Background:

• Iterative algorithms like maximum likelihood-expectation maximization (ML-EM) are standard for emission computed tomography (ECT) image reconstruction.
• ML-EM is sensitive to noise, leading to degraded reconstructions with increased iterations.

Purpose of the Study:

• To investigate a novel iterative algorithm for penalized-likelihood image reconstruction in ECT.
• To address the noise sensitivity and degradation issues of standard iterative algorithms.

Main Methods:

• Developed a new iterative algorithm incorporating fuzzy nonlinear anisotropic diffusion (AD) as a penalty function.
• Employed fuzzy reasoning to calculate diffusion coefficients, controlling diffusion strength via linguistic fuzzy rules.
• Utilized fuzzy set theory for uncertainty handling and nonlinear AD for noise reduction and edge preservation.

Main Results:

• The proposed algorithm converges to low-noise solutions even at high iteration counts, unlike ML-EM.
• Quantitative analysis demonstrated superior performance compared to ML-EM, OS-EM, Gaussian-MAP, MRP, and TV-EM.
• The method effectively removes noise while preserving crucial image edges.

Conclusions:

• The fuzzy nonlinear anisotropic diffusion-based iterative algorithm offers improved image reconstruction in ECT.
• This approach provides a robust alternative to existing methods, yielding high-quality images with reduced noise and preserved details.