Properties of Noise in Positron Emission Tomography Images Reconstructed with Filtered-Backprojection and Row-Action Maximum Likelihood Algorithm

Abstract

Noise levels observed in positron emission tomography (PET) images complicate their geometric interpretation. Post-processing techniques aimed at noise reduction may be employed to overcome this problem. The detailed characteristics of the noise affecting PET images are, however, often not well known. Typically, it is assumed that overall the noise may be characterized as Gaussian. Other PET-imaging-related studies have been specifically aimed at the reduction of noise represented by a Poisson or mixed Poisson + Gaussian model. The effectiveness of any approach to noise reduction greatly depends on a proper quantification of the characteristics of the noise present. This work examines the statistical properties of noise in PET images acquired with a GEMINI PET/CT scanner. Noise measurements have been performed with a cylindrical phantom injected with 11 C and well mixed to provide a uniform activity distribution. Images were acquired using standard clinical protocols and reconstructed with filtered-backprojection (FBP) and row-action maximum likelihood algorithm (RAMLA).Statistical properties of the acquired data were evaluated and compared to five noise models (Poisson, normal, negative binomial, log-normal, and gamma). Histograms of the experimental data were used to calculate cumulative distribution functions and produce maximum likelihood estimates for the parameters of the model distributions. Results obtained confirm the poor representation of both RAMLA-and FBP-reconstructed PET data by the Poisson distribution. We demonstrate that the noise in RAMLAreconstructed PET images is very well characterized by gamma distribution followed closely by normal distribution, while FBP produces comparable conformity with both normal and gamma statistics.

The Gamma Characteristic of Reconstructed PET Images: Implications for ROI Analysis

“…3 A and B. In the present study, the FBP images show no skewness supporting the Gaussian distribution consistent with previous studies [22]. By contrast, the IR images show skewness particularly in the early time frames (Fig.…”

“…Using quantile-quantile (Q-Q) plots and less formal statistical methods, Teymurazyan et al [22] showed that the noise in RAMLA-reconstructed PET images was well characterized by the Gamma distribution while FBP reconstructions produced comparable conformity with both normal and Gamma statistics. This study uses a more formal likelihood based approach to investigate evidence in support of the Gamma model in the EM and FBP reconstructed images.…”

“…Many factors impact on the noise characteristics of reconstructed PET images and might lead to a deviation from the Poisson distribution: most obviously scaling but attenuation and scatter corrections add further complexities. Barret et al [3] and Li [12] theoretically analyzed the noise properties/propagation of iterative reconstruction algorithms and suggested that the probability density function of the grey level at each pixel of the reconstructed images can be approximated by a log-normal distribution.Teymurazyan et al [22] reported measurements on a GEMINI PET/CT scanner comparing FBP and iterative EM reconstructions. While Gaussian characteristics were reasonable for FBP reconstructed images, the EM-reconstructed images were typically skewed.…”

Influence of Statistical Fluctuation on Reproducibility and Accuracy of SUVmax and SUVpeak: A Phantom Study

“…Positron emission itself is characterized statistically by a Poisson distribution (7). Even if the same object is scanned, the same image cannot be obtained because of the statistical fluctuation (8).…”

The Current Role of Image Compression Standards in Medical Imaging

“…Similarly, the spatial resolution in PET is typically between 3 to 5 mm and, thus, PET data sets are relatively small as well. Noise in SPECT and PET systems are due to several factors including the Poisson distributed noise during emission, electronic noise in the scanner's detection system, and the noise alterations due to the post-processing corrections and the image reconstruction process [24].…”