350 rub
Journal Information-measuring and Control Systems №2 for 2011 г.
Article in number:
Development of a way and method of creation of information-measuring intellectual system of diagnostics of a hepatites
Authors:
M. A. Ball, V. N. Meshcheryakov, V. V. Kavygin, S. V. Polozov
Abstract:
For help specialist advisable to use information and intellectual possibilities of the computer systems, in particular, computer expert systems. They use the different principles of the work. One of the approach of the decision given problems is a development of the intellectual systems for shaping the dependencies between unorganized the difficult formalized by input information and taken decisions. In medicine of such problem appear when stating diagnosis. The difficulty of the diagnostics viral hepetitis is concluded in need to take into account the big amount a factor. The Intellectual system like alive organism forms the skills, on the grounds of which and are taken the decisions. The skills are formed automatically, way recurrence-iteration education pro rata hearing mistake. Formed as a result of education weight factors similar conductivity synapse cortical neuron alive organism. For structured realization skill systems of the diagnostics hepetitis are stated some requirements, in particular, use the intellectual possibilities of the system by workman of the medical sphere must be intuitive comprehensible, adjusting the system must be for-give;pardon and reduce to minimum. The modules neural networks are arranged in the manner of dynamic libraries and fit in one directory with program-shell. In process of the functioning (when making skill systems) program shell searches for in the current directory of the library with module of the network (for this necessary simplest protocol, providing finding and identification of the modules skill systems). Control skill by system is realized from shell with use certain beforehand determined set function exported from module DLL. The functions realize the operations of low-level control system. The development skill systems was winnowed in several stages, gradually complicating set the problem. As a result, amount input factor was increased with 11 serological marker, before 58 factors, defining factors of the clinical examination, epidanamnesis, ul-trasonic examination and etc. Beside 600 medical cards Lipeckoy clinical infectious hospital were used For shaping the sample input data. The Results of the testing the functioning skill systems has shown that inaccuracy of the response to network on test example, does not exceed 10%, but amount it is correct classified diagnosis forms 94%.
Pages: 62-70
References
  1. Бала М. А., Кавыгин В.В., Полозов С. В. Обучаемая интеллектуальная система диагностики гепатитов // Новые биокибернетические и телемедицинские технологии XXIвека для диагностики и лечения заболеваний человека «НБИТТ-21». Междисциплинарная конференция с международным участием. 21-23 июня 2004 г, г. Петрозаводск.
  2. Мещеряков В. Н., Кавыгин В. В., Полозов С. В. Обучаемые навыковые системы: технического зрения, диагностики гепатитов, экологического мониторинга и выбора оборудования // Информационно-измерительные и управляющие системы, 2006, Т. 4, № 1-3.
  3. Кавыгин, В. В., Чесноков А. Т., Полозов С. В. Программная система «Навыковая система диагностики гепатитов «Серологические маркеры 1.0» // Зарегистр. в ВНТИЦ №50200601695 от 26.09.2006.