Skip to main content
. 2022 Dec;63(12):1887–1890. doi: 10.2967/jnumed.122.263902

TABLE 2.

Diagnostic Test Statistics for the Dichotomized SUVmax Result Versus Pathology

1.8 cutoff value for SUVmax 3.5 cutoff value for SUVmax
Reader Sensitivity Specificity PPV NPV Sensitivity Specificity PPV NPV
Expert H&N scan reader 1 0.471 (0.327, 0.623) 0.894 (0.862, 0.923) 0.268 (0.167, 0.381) 0.954 (0.931, 0.972) 0.300 (0.155, 0.459) 0.965 (0.942, 0.982) 0.412 (0.231, 0.611) 0.944 (0.919, 0.965)
Expert H&N scan reader 2 0.250 (0.109, 0.419) 0.900 (0.868, 0.929) 0.167 (0.070, 0.281) 0.938 (0.910, 0.962) 0.183 (0.062, 0.330) 0.967 (0.947, 0.983) 0.306 (0.116, 0.517) 0.937 (0.911, 0.961)
Combined central SUVmax 0.507 (0.356, 0.652) 0.851 (0.814, 0.884) 0.225 (0.142, 0.315) 0.953 (0.930, 0.972) 0.267 (0.135, 0.412) 0.970 (0.952, 0.986) 0.435 (0.243, 0.634) 0.939 (0.915, 0.961)
Site read 0.395 (0.250, 0.548) 0.903 (0.874, 0.930) 0.263 (0.154, 0.383) 0.945 (0.920, 0.966) 0.250 (0.119, 0.395) 0.972 (0.955, 0.987) 0.442 (0.235, 0.658) 0.937 (0.912, 0.959)

H&N = head and neck.

Data in parentheses are 95% CIs.