earticle

영어

This paper compares the performance of the combination of biomarkers to distinguish benign tumor from cancer and normal from cancer, from 21 urine biomarkers. Samples consist of 79 healthy women, 119 patients with benign tumor, and 137 patients with ovarian cancer. The concentrations of the 21 biomarkers were extracted using Luminex-PRA. The area under the curve (AUC) of ROC was evaluated to determine the optimum marker combination showing the best performance. The performance of the selected combination was confirmed with logistic regression. The highest AUC value of distinguishing benign tumor from cancer with a combination of two biomarkers and three biomarkers are 87.97% and 91.38%, respectively. And the highest AUC value of distinguishing normal from cancer with a combination of two biomarkers and three biomarkers are 87.39% and 90.68%, respectively. Interestingly, the benign-cancer classification shows a little higher performance than the normal-cancer classification.