Repeatability of SUV measurements in serial PET

Area of Science:

• Nuclear Medicine
• Medical Imaging Physics
• Quantitative PET Imaging

Background:

• Standardized uptake value (SUV) is crucial for monitoring cancer treatment response using FDG-PET/CT.
• Understanding the statistical uncertainty of SUV measurements is essential for accurate interpretation.

Purpose of the Study:

• To assess the reproducibility and statistical uncertainty of SUV max and SUV mean.
• To determine the minimum uncertainty threshold for physiological change detection in SUV.

Main Methods:

• Measurements were performed on a GE Discovery STE PET/CT scanner using a 68Ge phantom, an 18FDG phantom with varying target sizes, and eight cancer patients with healthy livers.
• Acquisition times ranged from 3 to 30 minutes for phantoms and included sequential 3-minute scans for patients.
• Repeatability coefficients and average percent changes in SUV max and SUV mean were calculated.

Main Results:

• For the 68Ge phantom, repeatability coefficients for SUV max and SUV mean decreased with longer acquisition times (3-30 min).
• In the 18FDG phantom, SUV max and SUV mean fluctuations decreased with increasing target diameter.
• Patient liver scans showed variability in SUV max and SUV mean repeatability, with overall ROI repeatability ranging from 27.9% to 41.1%.

Conclusions:

• SUV mean exhibits approximately half the statistical fluctuation of SUV max.
• For short scan durations (approx. 3 min) and small regions, SUV max and SUV mean have similar statistical fluctuations.
• Increased scanning time or larger target regions significantly reduce the statistical fluctuations of SUV mean compared to SUV max.