Artificial Inteligence Methods in the Problems of Drug Side-Effects

The paper has been devoted to the problems of prognosis of drag side-effects of chemotherapy application and it shows the possibility of effective use of artificial neural networks to this perpose. Based on meny years obsevations of chemotherapy patients have been gathered and elaborated the suitable training sets and testing sets for the neural networks. In the paper following cases has been taken into considerations: a) five kinds of chemotherapy programmes (schedules): fluorouracyl + leukovorin, with the application of platinum derivative, with the taxan application, with the irynotecan application and with the antracyclin application); b) 45 kinds of side–effects symptoms with proper types of scales.; c) five features describing a patient condition. The paper ends with conclussions describing practical uses of the created neural networks.