Skip to main content
. 2019 Jun 11;19:564. doi: 10.1186/s12885-019-5629-x

Table 6.

The calculated performance indices for different models and SUA for the testing group (n = 100)

Model / Method Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) AUC
ADNEX 88.0% 94.0% 93.6% 88.7% 0.972
(76.2–94.4) (83.8–97.9) (82.8–97.8) (77.4–94.7) (0.946–0.999)
LR2 72.0% 98.0% 97.3% 77.8% 0.969
(58.3–82.5) (89.5–99.9) (86.2–99.9) (66.1–86.3) (0.936–1.0)
RMI3 82.0% 88.0% 87.2% 83.0% 0.912
(69.2–90.2) (76.2–94.4) (74.8–94.0) (70.8–90.8) (0.854–0.970)
RMI4 84.0% 86.0% 85.7% 84.3% 0.932
(71.5–91.7) (73.8–93.0) (73.3–92.9) (72.0–91.8) (0.882–0.983)
SRrisk 82% 96.0% 95.3% 84.2% 0.976
(69.2–90.2) (86.5–98.9) (84.5–98.7) (72.6–91.5) (0.953–0.999)
SUA 92.0% 94.0% 93.9% 92.2% 0.930
(81.2–96.8) (83.8–97.9) (83.5–97.9) (81.5–96.9) (0.880–0.981)
Our model 96.0% 86.0% 87.3% 95.6% 0.977
(86.5–98.9) (73.8–93.0) (76.0–93.7) (85.2–98.8) (0.955–0.999)

95% CI 95% confidence intervals, ADNEX, LR2, RMI3, RMI4, SRrisk abbreviations for different models (details in the text), AUC area under the curve, NPV negative predictive value, PPV positive predictive value, RMI risk of malignancy index (model), SUA subjective ultrasound assessment