TABLE 4.

PET Data Analysis Characteristics of 19 Studies Reviewed

StudyParameter variableROI techniqueObserver
Schulte et al. (5)TBRROIs were individually defined, expressing maximum tumor uptake, excluding areas of lower uptake within tumor. Identical configuration at contralateral extremity was used to obtain TBR. In each case ROIs > 2.6 cm2.2, independent (blinded)
Franzius et al. (6)T/NTRectangular ROI was positioned around tumor activity in coronary slice with maximum tumor activity, with boundaries of ROI located just within apparent hypermetabolic zone.2, in consensus (blinded)
Smith et al. (7)Influx constant K, DURBSAROIs were manually drawn around each lesion. Maximum pixel value of DUR or influx constant K within ROI was recorded.2, in consensus (blinded)
Nair et al. (8)TBRIdentical ROIs were placed over tumor and contralateral normal limb.3, independent (not mentioned whether blinded)
BrĂ¼cher et al. (9)SUVmeanCircular ROI (1.5 cm in diameter) was manually placed in slice with maximum 18F-FDG uptake. SUVs were calculated using average activity values in ROI.Not mentioned
Ryu et al. (10)SUVmeanSUV of primary tumor was determined as mean value in 12-mm ROI positioned over area with highest activity within tumor as determined by visual analysis.2, independent (blinded)
Kitagawa et al. (11)SUVmeanRound ROIs (5 mm in diameter) were placed over area of highest 18F-FDG uptake in tumor on static images. SUV = tissue radioactivity concentration (Bq/mL)/injected dose (Bq) per body weight (g).3, independent (blinded)
Brink et al. (12)SUVmeanAverage activity values were determined in intratumoral ROI placed on slice with maximum activity concentration.2, independent (not mentioned whether blinded)
Chen et al. (13)SUVNot mentioned.1, not mentioned whether blinded
Wieder et al. (14)SUVmeanCircular ROIs (1.5 cm in diameter) were manually placed over all tumors at site of maximum 18F-FDG uptake on baseline scan. SUVs normalized to patient body weight were calculated from average activity values in ROI.Not mentioned
Song et al. (15)SUVmaxFor semiquantitative analysis of increased 18F-FDG uptake lesion, maximum SUV based on body weight was calculated.1
Cascini et al. (16)SUVmeanIrregular ROIs were semiautomatically drawn manually on transaxial planes using region-growing method that included pixels above threshold value (between 20% and 50% of maximum pixel value). Table 1: SUVmean.1, not mentioned whether blinded
Huang et al. (17)SUVmeanROIs were hand-drawn over tumor for calculation of SUV. ROIs were drawn to follow contours of elevated 18F-FDG activity, as compared with normal tissue, contralateral to tumor site.Not mentioned
Wieder et al. (18)SUVmeanROIs were manually placed over each primary tumor. Circular ROI of 1.5 cm (1.5 cm in diameter; 10 pixels) was placed on slice with maximum 18F-FDG uptake. SUVs were calculated using average activity values in ROI.Not mentioned
Iagaru et al. (19)SUVmaxROIs were placed around regions of increased 18F-FDG uptake for SUVmax determination.Not mentioned
Nishiyama et al. (20)SUVmaxSUV was defined as tissue concentration of 18F-FDG (kBq/mL) in structure delineated by ROI divided by activity injected per gram of body weight (kBq/g). ROI was placed over entire primary tumor. SUVmax of primary tumor was used.2, not mentioned whether independent or in consensus or whether blinded
Benz et al. (21)SUVmaxManual delineation of ROI on consecutive axial slices of CT scan was used. SUVmax and SUVmean were calculated.1, blinded
Smithers et al. (22)SUVmaxMaximum voxel activity in tumor was used for SUV quantification.1, not mentioned whether blinded
Ye et al. (23)SUVmaxROIs were individually defined for each patient on transverse sections of PET images. SUVmax was measured.2, independent (blinded)
  • TBR: tumor-to-background ratio; T/NT: tumor-to-nontumor ratio; DUR: dose uptake ratio; DURBSA = dose uptake ratio body surface area; SUVmax = maximum standardized uptake value; SUVmean = mean standardized uptake value.