Myocardial viability is certainly an important area for clinical investigation. My major concerns with previous work has been in the variability of 'definition' and lack of 'gold standard'. What would your definition be and what model were you interested in studying?

I think the problem is does adding a specific layer or methodology and the definition of myocardial viability based upon that constuct improve the reliability of the model to make reasonable predictions for clinical utilization. All techniques have a sensitivity and specificity for basing their results on.

My question is "does adding 2 or 3 methodologies together improve the likilihood of a more accurate model?"

I understand your question about combining methods, but a clear definition of "viability" would still be required. Functional recovery of the myocardium after coronary revascularization is but one aspect of viability (and I believe this is too limited). Arrhythmogenity, myocardial response to exercise, and possibly other parameters should also be considered. Which methodologies did you hope to combine? Were you interested in a prospective study? Human or animal?

I believe simply labeling something "viable" is inadequate. It has to be evidence based to justify the idea of what is best to define to define myocardial viability using the constraint of a model. However, technology would vary in many parts of the world so some models would be more accurate than other. As you describe, the model could include functional recovery of the myocardium before or after coronary revascularization, ECG, echocardiogram, arrhythmogenity, myocardial response to exercise, image evaluation with nuclear or MRI imaging.

Essentially, a multivariate model would essentially describe what is myocardial

viability and would give a a statistical answer to this semantic enigma.

I would suggest using a retrospective study using data already available in population groups with cardiac disease in humans.