Phantom and clinical evaluation of Block Sequential Regularized Expectation Maximization (BSREM) reconstruction...

Area of Science:

• Nuclear Medicine
• Medical Imaging Physics

Background:

• Positron Emission Tomography (PET) imaging is crucial for cancer diagnosis and monitoring.
• Image reconstruction algorithms like Ordered Subset Expectation Maximization (OSEM) and Block Sequential Regularized Expectation Maximization (BSREM) impact image quality.
• Determining optimal parameters for BSREM is essential for accurate lesion detection and characterization.

Purpose of the Study:

• To investigate the impact of varying β-values in the BSREM algorithm on lesion reconstruction.
• To identify optimal β-values for different lesion sizes in 68Ga-PSMA PET/CT imaging.
• To compare BSREM performance with OSEM for phantom and clinical prostate cancer data.

Main Methods:

• Utilized a National Electrical Manufacturers Association phantom and 15 prostate cancer patients scanned with 68Ga-PSMA PET/CT.
• Reconstructed images using OSEM and BSREM with a range of β-values.
• Quantified image quality metrics including background variability, contrast recovery, signal-to-noise ratio, and signal-to-background ratio (SBR).

Main Results:

• BSREM with lower β-values improved contrast recovery and signal-to-noise ratio compared to OSEM.
• Optimal β-values for reconstruction were found to be inversely related to lesion size.
• Specific β-values (400 for small, 600/500 for large lesions) demonstrated superior performance in phantom and clinical data.

Conclusions:

• The choice of β-value in BSREM significantly affects the reconstruction of lesions of varying sizes.
• Optimal β-values of 400, 600, and 500 are recommended for small, and large lesions in phantom and clinical 68Ga-PSMA PET/CT studies, respectively.
• This parameter optimization enhances diagnostic accuracy in prostate cancer imaging.