A new method for PET image reconstruction using Fourier-Wavelet moment

Area of Science:

• Medical Imaging
• Signal Processing
• Computational Science

Background:

• Positron Emission Tomography (PET) imaging requires robust reconstruction algorithms.
• Conventional methods often rely on regularization, which can introduce artifacts or limit resolution.
• There is a need for non-regularization techniques that enhance reconstruction quality and efficiency.

Purpose of the Study:

• To introduce a novel non-regularization method for PET image reconstruction.
• To leverage a feature-based approach utilizing a Fourier-Wavelet basis.
• To improve reconstruction quality and computational efficiency in PET.

Main Methods:

• A feature-based reconstruction method employing a Fourier-Wavelet basis.
• Calculation of Fourier-Wavelet moments (FWM) from PET measurements.
• Utilization of an iterative approach with a row-action (RA)-like algorithm for accelerated convergence.
• Exploitation of the rotation invariance property of the Fourier-Wavelet basis to reduce computational load.

Main Results:

• The proposed method achieves good reconstruction quality in experimental comparisons.
• It demonstrates superior performance compared to conventional Maximum A Posteriori (MAP) methods.
• The use of Fourier-Wavelet basis and RA-like algorithm enhances efficiency and convergence.

Conclusions:

• The developed non-regularization method offers a promising alternative for PET reconstruction.
• The Fourier-Wavelet basis provides advantages in terms of computational cost and reconstruction quality.
• This approach represents a significant advancement in PET imaging technology.