Variation and repeatability of measured standardized uptake values depending on actual values: a phantom study

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Quantitative Analysis

Background:

• Standardized uptake values (SUVs) are crucial quantitative imaging biomarkers in Positron Emission Tomography (PET).
• The variability and repeatability of SUVs across different value magnitudes are not well understood.
• Understanding SUV variation is essential for accurate interpretation of PET scans.

Purpose of the Study:

• To investigate how the magnitude of Standardized Uptake Values (SUVs) affects their variation and repeatability.
• To test the hypothesis that lower SUVs exhibit greater variation compared to higher SUVs.
• To evaluate the reliability of SUV measurements in Positron Emission Tomography (PET) imaging.

Main Methods:

• Experimental PET scans were conducted using a NEMA IEC body phantom with adjustable F-18 concentrations to simulate SUVs of 2.0, 4.0, and 8.0.
• PET data were acquired over 4 hours with reconstructions every 2 minutes.
• SUVmax and SUVpeak values were recorded, and relative SUVs were calculated for Bland-Altman analysis.

Main Results:

• Measured SUVs exhibited significant variation, with the lowest actual SUV (2.0) showing the largest variation and the highest SUV (8.0) showing the smallest.
• Relative SUVs demonstrated substantially greater variation at lower values.
• Bland-Altman plots confirmed that measurement variation decreased as the measured SUV value increased.

Conclusions:

• The study demonstrates that measured SUV values have inherent variability that decreases with increasing magnitude.
• Clinicians must consider the potential inaccuracy of low SUV measurements in daily practice.
• The findings are critical for multi-institutional PET studies where consistent SUV interpretation is vital.