FDG-PET standardized uptake values in normal anatomical structures using iterative reconstruction segmented...

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiochemistry

Background:

• Filtered back-projection (FBP) is standard for PET image reconstruction but often yields noisy images.
• Iterative reconstruction segmented attenuation correction (IRSAC) enhances image quality without compromising resolution.
• Standardized uptake value (SUV) is crucial for quantifying [fluorine-18]fluoro-2-deoxy-D-glucose (FDG) uptake in PET.

Purpose of the Study:

• To establish reference SUV values for normal anatomical structures using both FBP and IRSAC.
• To compare SUV measurements obtained from FBP and IRSAC reconstructed PET images.
• To evaluate the impact of different transmission scan filters on SUV quantification.

Main Methods:

• Retrospective analysis of 20 whole-body PET scans from non-small cell lung cancer patients.
• Image reconstruction using both IRSAC and FBP algorithms.
• Phantom study to assess filter effects on SUV; calculation of SUVmax and SUVavg in 28 normal structures and one pathological site.

Main Results:

• IRSAC with a Gaussian filter yielded SUVavg and SUVmax values 20% and 8% higher than FBP, respectively.
• A non-quantitative filter for transmission scans led to a 20% underestimation of SUV.
• IRSAC identified most anatomical structures, showing a strong overall correlation with FBP SUVs.

Conclusions:

• IRSAC with a Gaussian filter offers more accurate SUV quantification than FBP with a non-quantitative filter.
• IRSAC improves the accuracy of quantitative PET imaging, particularly for lower SUV values.
• Reference SUV values were established for normal anatomical structures using both reconstruction methods.