“…With respect to the algorithmic implementation, the fact that the criterion is enclosed into a black-box statistical model makes it unattractive for a medical professional, as long as the nature of the problem remains in the SVM formula. This subject will be studied in the companion article also in this issue [16].…”