Concept And Interpretation Of ROC Curve

Question 1

The concept of a ROC curve is based?

Answer 1

the notion of a “separator” (or decision) variable

Question 2

What happens if one changes the criterion or cut off?

Answer 2

frequencies of positive and negative results of the diagnostic test will vary if one changes the “criterion” or “cut-off” for positivity on the decision axis.

Question 3

Latent or unobservable variable?

Answer 3

Where the results of a diagnostic system are assessed based on subjective judgement, the decision scale is only “implicit”. Such a decision variable is often called a “latent” or unobservable variable.

Question 4

ROC curve lying on the diagonal line?

Answer 4

reflects the performance of a diagnostic test that is no better than chance level, i.e. a test which yields the positive or negative results unrelated to the true disease status.

Question 5

The AUC?

Answer 5

is effective and combined measure of sensitivity and specificity that describes the inherent validity of diagnostic tests.

Question 6

What is the example of real data breast cancer?

Answer 6

As an example of real data of breast cancer study that was reported recently by Hajian-Tilaki et al. (12), the calculation of sensitivity and specificity of various cut-off values of body mass index (BMI) for predicting of breast cancer are given in table 2 and shows that sensitivity and specificity has an inverse=käänteinen relationship

This curve and the corresponding AUC show that BMI as a biomarker has predictive ability to discriminate breast cancer from normal subjects.