Suppose the following ROC curve is obtained for a prostate cancer screening test, where a high score on the test indicates higher probability of having prostate cancer. Using a low threshold will detect fewer true cases, but also have fewer false positives. Setting a higher threshold detects more cases, but at the cost of more false positives. Based on this ROC curve, what is the maximum number out of each 100 true cases that is it possible to detect without having any false positives? Sensitivity 1.0 0.8 0.6 0.4 zo 0.0 .0 • 10 .30 •50 1.0 Highlight the correct answer. • 60 • 90 • 100 0.5 Specificity 00

Suppose the following ROC curve is obtained for a prostate cancer screening test, where a high score on the test indicates higher probability of having prostate cancer. Using a low threshold will detect fewer true cases, but also have fewer false positives. Setting a higher threshold detects more cases, but at the cost of more false positives. Based on this ROC curve, what is the maximum number out of each 100 true cases that is it possible to detect without having any false positives? Sensitivity 1.0 0.8 0.6 0.4 zo 0.0 .0 • 10 .30 •50 1.0 Highlight the correct answer. • 60 • 90 • 100 0.5 Specificity 00