Objective:

To investigate the diagnostic consistency between physicians in differentiating malignant from benign ovarian tumors by using IOTA LR2 model.

Methods:

The ultrasonographic images of 876 ovarian tumors were retrospectively analyzed by two physicians with different seniorities in the absence of clinical and pathological data. The IOTA LR2 model was used to evaluate the ultrasonographic images. The results were compared with the pathological results to determine the diagnostic efficacy. Kappa test was used to determine the consisten

cy between two physicians.

Results:

Among 876 ovarian tumors

837 cases were benign and 39 were malignant. The sensitivity and specificity of the two physicians were 76.9%

67.0% and 71.8%

76.0%

respectively

and the area under the receiver operating characteristic (ROC) curve was 0.8680.040 and 0.8470.043

respectively. Kappa was 0.659 between the two physicians (

P

0.01).

Conclusion:

IOTA LR2 model has a high diagnostic value in differentiating malignant from benign ovarian tumors

and the diagnostic consistency between physicians is good.