[Multi-center study of inter-scanner difference in brain positron emission tomography]

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Radiochemistry

Background:

• Positron emission tomography (PET) is vital for brain imaging using tracers like (18)F-FDG and (11)C-PiB.
• Scanner variability can impact the accuracy and reliability of quantitative PET measurements.
• Multicenter studies require consistent imaging performance across different PET systems.

Purpose of the Study:

• To evaluate scanner dependence in brain PET imaging.
• To assess the impact of phantom characteristics and external radioactivity on image quality.
• To investigate methods for improving axial uniformity in PET scans.

Main Methods:

• Phantom studies using cylindrical and 3D brain phantoms with nine different PET scanners.
• Evaluation of axial uniformity using regions of interest.
• Assessment of scatter and random coefficients influenced by external radioactivity.
• Application of a novel dead-time correction method.

Main Results:

• Significant scanner dependence observed for (18)F-FDG and (11)C-PiB uptake values.
• Standardized uptake values (SUV) were underestimated by some scanners.
• Axial uniformity was degraded by radioactivity outside the axial field of view (AFOV) in certain scanners.
• Dead-time correction improved uniformity in the cylindrical phantom but not the 3D brain phantom.

Conclusions:

• PET scanner performance varies, impacting quantitative accuracy.
• External radioactivity can negatively affect axial uniformity.
• A new dead-time correction method shows potential but requires further refinement for complex phantoms.
• Standardized quality control of PET scanners is essential for multicenter research.