FDG PET is an important tool for assessing tumours and monitoring therapeutic response. The most commonly used quantitative parameters are based upon the standardized uptake value (SUV), often evaluated as the maximum, SUVmax. We propose a new parameter based upon generating a histogram of SUV values for the whole tumour and fitting a curve to obtain the peak of this distribution, SUVpeak. The applicability and robustness of SUVpeak have been evaluated for a range of lesions in phantom experiments. Reproducibility has been assessed on repeat images in a series of patients. SUVpeak is not significantly affected by histogram bin width (0.1 recommended), fitting function (Gaussian plus linear background recommended), lower limit of fitting range, and count density, with standard deviations of 0.5–1.0%. As might be expected, there is a lower volume limit of approximately 5 ml below which SUVpeak is indeterminable, making the technique more applicable to assessment of larger lesions. Reproducibility of SUVmax in patients was good with a correlation coefficient of R2=0.948 but even better for SUVpeak with R2=0.985. We have developed a new quantitative parameter of FDG uptake, SUVpeak, which is more reproducible than SUVmax and has the potential to be a more sensitive tool in assessing therapy response.
No takes yet. Share an insight, caveat, or question.
Julyan et al. (2004) studied this question.