Investigation of the Effect of Bayesian Penalized Likelihood (BPL) Reconstruction Algorithms on The FDG PET Volumetric And Heterogeneity Features: A Phantom Study


1 Tabriz University of Medical Sciences*

2 Department of Medical Physics and BioMedical Engineering Faculty of Medicine, Tehran University of Medical Sciences

3 Department of Nuclear Medicine Imam Khomeini Hospital Complex, Tehran University of Medical Sciences

4 Department of NeuroScience and Addiction Studies School of Advanced Technologies in Medicine, Tehran University of Medical Sciences, Tehran, Iran

5 Department of Medical Physics Faculty of Medicine, Tabriz University of Medical Sciences

6 Department of Medical Physics Iran University of Medical Sciences



The purpose of this study is to investigate the effect of Bayesian penalized likelihood (BPL) image reconstruction algorithms with different noise-controlling factor (β-values) on the FDG PET quantitative volumetric and heterogeneity features of metabolic volume (MTV) and total glycolysis (TLG).
Materials and Methods:
The standard National Electrical Manufacturers Association (NEMA) image quality phantom that consisted of six spheres (inner diameters of 37, 28, 22, 17, 13, and 10 mm) filed with 18F-FDG solution were used for this study. The phantom was prepared with a lesion-to-background ratio
(LBR) of 4:1. All images were reconstructed using OSEM+ resolution recovery and BPL using β-values of 100, 200, 300, 400, 500. The metabolic tumor volume (MTV) and total lesion glycolysis (TLG) were measured for all spheres. The relative percent error (ΔMTV and ΔTLG) with respect to true values, mean difference, standard deviation, and inter-reconstruction variabilities were quantified and evaluated.
The mean ΔMTV showed statistically significantchangesbetweenBPLreconstruction method in different β-values and OSEM, at LBR4 were (-1.83%, -13.80%, -0.60%, -0.34%
and 0.03%) statistically significant. There was also a significant difference in the, mean ΔTLG between BPL reconstruction methods and OSEM (-4.15, -2.67, -2.21, -1.81 and 1.51,
respectively) was observed (p < 0.05).
In this study, we indicated that both volumetric and heterogeneity features of ΔMTV and ΔTLG are sensitive to reconstruction algorithms. The BPL algorithm with different noise controlling factors (β value) provided different MTV and TLG values. We showed that significant differences of MTV and TLG can be obtained in low-level Beta values.