Fig. 2

Development of the ovarian cancer detection model (model 2). a Work flow of patients for development of prediction model 2. Serum samples were obtained from 1402 subjects, including 428 patients with ovarian tumors, 859 with other cancers, and 115 non-cancer controls (from non-cancer control B). The sample set was divided into two groups, the discovery set and validation set. b ROC curves for detecting cancer patients using the miRNAs selected for prediction model 2. c Diagnostic performance of the ten selected miRNAs in the discovery set and validation set. d Diagnostic index using prediction model 2 in the validation set (ovarian carcinoma, 160; breast carcinoma, 100; colorectal adenocarcinoma, 100; esophageal squamous cell carcinoma, 73; gastric adenocarcinoma, 100; hepatocellular carcinoma, 66; lung carcinoma, 100; pancreatic ductal adenocarcinoma, 100; sarcoma, 100; non-cancer, 100; non-epithelial ovarian cancer, 13; borderline ovarian tumor, 66; and benign ovarian tumor, 29). Each diagnostic accuracy (%) is indicated. The p values were calculated by χ² test