What is SUV and what affects it?
SUV (Standardised Uptake Value) is a semi-quantitative measure used in PET imaging to estimate tracer uptake within a region, normalised for administered activity and patient size. It is most commonly used in FDG PET to assess metabolic activity.
SUV is a normalised measure of PET tracer uptake that depends on scanner calibration, correction accuracy, reconstruction parameters, and biological variables.
SUV is calculated by dividing the measured activity concentration in tissue by the injected activity per unit body mass (or alternative normalisation factor). Although widely used, SUV is influenced by multiple biological and technical factors and is not a direct measure of metabolic rate.
Understanding the physics
SUV represents a normalised activity concentration. In PET imaging the SUV is calculated as the ratio of radioactivity concentration in a tissue region to the injected dose adjusted for body weight.
SUV = Activity concentration in tissue / (injected activity / body weight)
This calculation assumes uniform tracer distribution proportional to body mass. The activity concentration in tissue is that measured by the scanner at a particular region of interest (ROI) at a specific time point. The injected activity is decay-corrected for scan time. However, several factors influence the measured SUV.
From a physics perspective, SUV depends on:
Accurate scanner calibration
Correct attenuation and scatter correction
Adequate count statistics
Reconstruction parameters
Partial volume effects
Small lesions may have underestimated SUV due to spatial blurring. Excessive image smoothing reduces peak SUV values. Differences in reconstruction algorithm or iteration number can change reported SUVs.
From a biological perspective, SUV is influenced by:
Time between injection and imaging
Blood glucose level (in FDG imaging)
Patient body composition
Tracer pharmacokinetics
SUV can be reported as SUVmax (maximum voxel value), SUVmean (average within a region), or SUVpeak (average within a small fixed region around the maximum).
SUVmax is less operator-dependent but more sensitive to noise. SUVmean is more stable but depends on region-of-interest definition.
SUV is semi-quantitative because it does not fully account for dynamic tracer distribution or kinetic modelling.
Where this matters clinically
SUV is widely used in oncology for tumour characterisation and treatment response assessment. Understanding the physical and biological factors affecting SUV prevents overinterpretation of small changes between scans.
Consistency in acquisition timing, reconstruction parameters, and patient preparation is essential for reliable comparison.