Noise Considerations for PET Quantification Using Maximum and Peak Standardized Uptake Value

Abstract

In tumor response monitoring studies with 18F-FDG PET, maximum standardized uptake value (SUVmax) is commonly applied as a quantitative metric. Although it has several advantages due to its simplicity of determination, concerns about the influence of image noise on single-pixel SUVmax persist. In this study, we measured aspects of bias and reproducibility associated with SUVmax and the closely related peak SUV (SUVpeak) using real patient data to provide a realistic noise context. Methods List-mode 3-dimensional PET data were acquired for 15 min over a single bed position in twenty 18F-FDG oncology patients. For each patient, data were sorted so as to form 2 sets of images: respiration-gated images such that each image had statistical quality comparable to a 3 min/bed position scan, and 5 statistically independent (ungated) images of different durations (1, 2, 3, 4, and 5 min). Tumor SUVmax and SUVpeak (12-mm-diameter spheric region of interest positioned so as to maximize the enclosed average) were analyzed in terms of reproducibility and bias. The component of reproducibility due to statistical noise (independent of physiologic and other variables) was measured using paired SUVs from 2 comparable respiration-gated images. Bias was measured as a function of scan duration. Results Replicate tumor SUV measurements had a within-patient SD of 5.6% ± 0.9% for SUVmax and 2.5% ± 0.4% for SUVpeak. SUVmax had average positive biases of 30%, 18%, 12%, 4%, and 5% for the 1-, 2-, 3-, 4-, and 5-min images, respectively. SUVpeak was also biased but to a lesser extent: 11%, 8%, 5%, 1%, and 4% for the 1-, 2-, 3-, 4-, and 5-min images, respectively. Conclusion The advantages of SUVmax are best exploited when PET images have a high statistical quality. For images with noise properties typically associated with clinical whole-body studies, SUVpeak provides a slightly more robust alternative for assessing the most metabolically active region of tumor.

Identification of Biomarkers Including 18FDG-PET/CT for Early Prediction of Response to Neoadjuvant Chemotherapy in Triple-Negative Breast Cancer

“…Because SUV max and SUV peak have both demonstrated lower interobserver variability than other quantitative PET parameters (36), SUV peak could also has been used, as suggested by Wahl and colleagues (37). Indeed, SUV peak is less sensitive than SUV max to the image noise and voxel size (38). However, the region of interest (ROI peak ) used to measure SUV peak is not uniquely defined (39,40).…”

The Relationship Between Estrogen Receptor, Progesterone Receptor and Human Epidermal Growth Factor Receptor 2 Expression of Breast Cancer and the Retention Index in Dual Phase 18F-FDG PET/CT

“…It was a predictable consequence, because SUVmax would be more expected to reflect the metabolic status of the tumor than SUVmean [33].…”

“…Because SUVmax is determined by a single pixel with the highest SUV [43], it could also be effected by its [33] reported that SUVmax was much more biased by noise properties than multi-pixel summarized SUV. In the current study, the PET images were acquired for 3.5 min for each bed, and therefore, the average positive biases of SUVmax and SUVmean are expected to be about 10 and 5 %, respectively.…”

Optimisation and harmonisation: two sides of the same coin?

“…Consequently, SUV max has become the de facto standard for quantitative assessment of tracer uptake. Despite the reported upward bias in SUV max [24,25], which becomes even more pronounced with new PET technologies ( Supplementary Fig. 1 in [22]), it has some attractive features.…”

“…SUV peak is based on a spherical volume of interest having a volume of ∼1 ml, positioned over the lesion such to obtain the highest average value. By averaging over 1 ml volume the upward bias as a function of image noise seen with SUV max may be avoided [25]. At the same time SUV peak may be less sensitive to underlying image characteristics as the dimensions of the peak volume of interest act as a smoothing filter [25].…”