ROC and AUC with (Sensitivity vs Specificity vs Accuracy)

ROC Receiver Operator Characteristics are used to measure properties of diagnostic tests by plotting the sensitivity vs one minus the specificity. A helpful summary of ROC studies is the Area Under the Curve (AUC). These are superior metrics to accuracy for diagnostic tests as the are less sensitive to the prevalence of the abnormalities in the population.

[@BadlndsBob]

I don't do radiology, but this presentation is very helpful for some other subjects, i.e., other lab tests.

[@HowRadiologyWorks]

Great 👍, yeah you’re right this is general for any test procedure. Please share with others in your lab 😉

[@anetter.gastelum7154]

6:39 In which program can I make the graph of the decision parameters? Thank you very much for the video!!

[@HowRadiologyWorks]

This was written in python but any language that does plotting could be used like Matlab

[@IllBeHonest1980]

if the disease is there and it predicts it is. Isn't that a True positive

[@HowRadiologyWorks]

Yes in my language that was True Abnormal and Observed Abnormal