diagnostic testing

Medical Testing

When evaluating a diagnostic test, one typical constructs a 2 x 2 table and calculates the specificity and sensitivity (true and false negatives, true and false positives); I believe this method is based on Bayes' theorem.

I’d like to evaluate diagnostic tests where the result is not + or – but rather a range of possible results like tumor size. If one test is better able to find smaller tumors (which may be more treatable) then that test may be of greater utility than one that can only identify larger tumors.

I assume the tumor size would be like a weighting factor.

Any thoughts or suggestions to help direct me?