Skip to main content
. 2022 Jan 13;156(3):559–567. doi: 10.1007/s11060-021-03936-y

Table 2.

ROC analysis of diagnostic accuracy of semi-quantitative imaging markers

Threshold value* Sens Spec pLR nLR AUC
SUVmax 5.82 0.70 0.92 8.26 0.33 0.88 (0.75, 1.00)
SUVpeak 5.00 0.70 0.97 20.65 0.31 0.88 (0.75, 1.00)
SULmax 8.83 0.67 0.90 6.56 0.37 0.86 (0.72, 0.99)
SULpeak 7.58 0.67 0.98 39.33 0.34 0.86 (0.72, 1.00)
Delayed SUVmax 2.53 1.00 0.56 2.29 0.00 0.81 (0.53, 1.09)
Delayed SUVpeak 2.08 1.00 0.63 2.67 0.00 0.85 (0.61, 1.10)
Delayed SULmax 3.42 1.00 0.50 2.00 0.00 0.69 (0.28, 1.10)
Delayed SULpeak 2.81 1.00 0.63 2.67 0.00 0.78 (0.44, 1.12)
TLmax 1.66 0.90 0.86 6.53 0.12 0.92 (0.84, 1.01)
TLmean 2.31 0.90 0.79 4.35 0.13 0.91 (0.81, 1.00)

Table showing ROC analysis of diagnostic accuracy of semi-quantitative imaging markers. *Optimal threshold obtained through Youden’s method. AUC area under the receiver operating curve; nLR negative likelihood ratio; pLR positive likelihood ratio; sens sensitivity; spec specificity; SUV standard uptake value; SUL standard uptake value adjusted for lean body mass; TLmax tumor-to-liver maximal ratio; TLmean tumor-to-liver mean ratio