350 rub
Journal Technologies of Living Systems №2 for 2009 г.
Article in number:
Forecasting of gallbladder cholesteroz development
Keywords:
cholelithiasis
a chronic cholecystitis
cholesterosis
identification
forecasting
the statistical methods
the discriminantal analysis
significant parameters
Authors:
V.B. Lifshits, T.I. Bouldakova, N.V. Ekimova, E.V. Ignatievna, S.I. Suyatinov
Abstract:
Problems of gallbladder cholesterosis development at healthy people and at the patients, suffering by a chronic cholecystitis, are considered. Diagnostics of disease is connected to the certain difficulties caused by absence of precisely developed clinical picture of gallbladder cholesterosis and criteria of laboratory and tool diagnostics.
We surveyed 91 patients (54 men and 37 women), middle age 50,06,2 years, for the decision of problems of forecasting. Depending on disease, patients have been divided into 3 groups: chronic cholecystitis during an aggravation; chronic cholecystitis during remission; cholesterosis. Also the control group has been surveyed. It included 25 practically healthy people. For the decision of tasks and gathering of diagnosing parameters it has been carried out clinico-anamnestic inspection, ultrasonic research of a bilious bubble on devices DP - 9900Plus / Mindray (CPR); SDU - 500C "Sreimadzu" (Japan) with frequency 3,5 MHz. Verification chronic cholecystitis and cholesterosis was achieved on the basis of classical clinical, sonographic and the laboratory data.
For the decision of a problem of identification and forecasting of diseases development two-level processing of the medical data has been realized. The purpose of preprocessing consist in the analysis of the saved up data, reduction of dimension of a task, a choice prognostic parameters. It was carried out by means of spreadsheet MS Excel. Correlation and cluster analysis have been executed with the help of software product STATISTICA. Correlation interrelations between clinical, sonographic and biochemical criteria have been established at chronic cholecystitis, cholesterosis and at healthy people.
At the second level the deeper data processing is carried out, the latent nonlinear interrelations between predictive parameters and probability of development of disease are revealed. The discriminantal analysis was applied to a rating of probability of cholesterosis development. The parameters which define physiological features of the person and controlled parameters (for example, smoking, a diet, an index of weight, etc.) were taken into account.
Three tasks of forecasting of cholesterosis development have been solved: a rating of probability of transition of a chronic cholecystitis (remission) in cholesterosis; a rating of probability of transition of a chronic cholecystitis (aggravation) in cholesterosis; a rating of probability of cholesterosis development at healthy people.
During the discriminantal analysis the most significant parameters for classification of functional statuses of patients have been revealed. Besides the discriminantal analysis has allowed to estimate adequacy of initial classification of tested people depending on forecasting parameters. The calculated coefficients allow to determine a concrete kind of discriminantal functions for a solved task and to construct mathematical models of forecasting
Pages: 52-59
References
- Лейшнер У. Практическое руководство по заболеваниям желчных путей. М.: Геотар-Медиа. 2001.
- Минушкин О.Н., Прописнова Е.П. Холестероз желчного пузыря (обзор) // Кремлевская медицина. 2000. №1. С. 55 - 57.
- Измайлова Т. Ф., Иванченкова Р. А. Холестероз желчного пузыря // Российский журнал гастроэнтерологии, гепатологии и колопроктологии. 1994. №4. С. 20 - 25.
- Василенко В.Х. Врачебный прогноз. Душанбе: Дониш. 1982.
- Медик В.А., Токмачев М.С. Математическая статистика в медицине. М.: Финансы и статистика. 2007.
- Гайдышев И. Анализ и обработка данных: специальный справочник. СПб.: Питер. 2001.
- Ким Дж.О., Мюллер Ч.У., Клекка У.Р. и др. Дискриминантный анализ // Факторный, дискриминантный и кластерный анализ. М.: Финансы и статистика. 1989.