A fast reconstruction algorthm for stationary positron emission tomography based on a modified em algorithm

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Computational Science

Background:

• Positron Emission Tomography (PET) imaging relies on iterative reconstruction algorithms.
• Standard expectation-maximization (EM) algorithms can be slow and sensitive to image characteristics.

Purpose of the Study:

• To develop a more efficient iterative reconstruction method for PET.
• To improve convergence speed and spatial resolution in PET image reconstruction.

Main Methods:

• An enhanced EM algorithm incorporating high-frequency components of projection data.
• Iterative correction of image density for uniform efficiency and flat frequency response.
• Simulation studies on scanning and stationary PET systems with sparse sampling.

Main Results:

• The enhanced algorithm achieves satisfactory PET images in 2-3 iterations.
• Significantly improved uniformity of spatial resolution compared to standard EM.
• Convergence speed is less sensitive to image patterns and matrix size.

Conclusions:

• The novel PET reconstruction algorithm offers faster computation and improved image quality.
• This method holds promise for reducing computational demands in PET imaging.
• Nonnegativity of reconstructed images is consistently preserved.