Использование нейросетевых классификаторов в прогнозировании тяжести течения и вероятности летального исхода в зависимости от этиологии острого панкреатита

Е. М. Лобынцева, Alexey L. Loktionov, Александр Иванович Конопля et al. · Вестник новых медицинских технологий · 2012

The present work is devoted consideration of problems of diagnostics of a pancreatitis, using neural network the approach to the analysis of real results of inspection of the patients spent to the first days of stay in a hospital, based on use self-training neural networks structures. From the practical point of view recognition of various forms of a pancreatitis, should is spent as much as possible productively, to short terms and on the basis of application of whenever possible limited number of routine methods of the tool and laboratory research which quantity defined by the standard nevertheless is great enough. The basic advantage of the given approach high degree of confidence of forecasting of the right diagnosis at the patient, using the minimum quantity of methods of tool and laboratory inspection.

Cited by 0 works.

Authors