Medical decisions are hard. One of the reasons for this difficulty is that they often have to be made on the basis of insufficient and uncertain information. Furthermore, the outcome of the decision process has far-reaching implications on the well-being or even the very lives of patients. There is a substantial body of empirical evidence that shows that human decision-making performance is far from optimal. Furthermore, its quality decreases with the complexity of problems, time pressure, and high stakes. Therefore, given the increasing diversity of treatments and drugs, the accumulating body of knowledge about human health, the availability of diagnostic tests, including genetic tests, medical specialists often need to be assisted in their decision-making. At the same time, increasing costs of medical care and amounting risk of litigations following medical errors increase the pressures on the medical community to become cost-effective and to make fewer errors. Given all this, despite the fact that the medical community is rather conservative and resists technological assistance, computer support for medical decision-making is an inevitable fact. This is true in a number of aspects of medical care, including its initial phase when a physician has to come up with a preliminary prognosis or diagnosis and specify the possible directions for the patient's treatment.