Skip to main content
. 2019 Oct 12;146(1):237–244. doi: 10.1007/s00432-019-03054-3

Table 2.

Relationship between CT/MR imaging and 18F-FDG PET/CT results and neck histopathology findings

Imaging No. patients Percentage (95% confidence interval)
TP FP FN TN Sensitivity Specificity Accuracy PPV NPV
Patients (n = 178)
 CT/MR 15 15 27 121 35.7 (21.6–52.0) 89.0 (82.5–93.7) 76.4 (69.5–82.4) 50.0 (34.8–65.2) 81.8 (78.0–85.0)
 PET/CT 29 30 13 106 69.1 (52.9–82.4) 77.9 (70.0–84.6) 75.8 (68.9–81.9) 49.2 (39.9–58.5) 89.1 (83.7–92.8)
 P value* 0.001 0.003 1.000
Neck sides (n = 199)
 CT/MR 16 16 28 139 36.4 (22.4–52.2) 89.7 (83.8–94.0) 77.9 (71.5–83.5) 50.0 (35.3–64.7) 83.2 (79.8–86.2)
 PET/CT 31 33 13 122 70.5 (54.8–83.2) 78.7 (71.4–84.9) 76.9 (70.4–82.6) 48.4 (39.6–57.3) 90.4 (85.5–93.7)
 P value < 0.001 0.001 0.880
Neck levels (n = 678)
 CT/MR 17 20 41 600 29.3 (18.1–42.7) 96.8 (95.1–98.0) 91.0 (88.6–93.1) 46.0 (32.1–60.5) 93.6 (92.5–94.5)
 PET/CT 36 67 22 553 62.1 (48.4–74.5) 89.2 (86.5–91.5) 86.9 (84.1–89.3) 35.0 (28.4–42.1) 96.2 (94.8–97.2)
 P value < 0.001 < 0.001 0.003

FN false negative, FP false positive, NPV negative predictive value, PPV positive predictive value, TN true negative, TP true positive

*McNemar test

Logistic regression using GEE (Generalized Estimating Equations) that accounted for the clustering of observations within patients