Heuristic approach to the classification of postpartum endometritis and its forms

Тhe work is dedicated to the development of a method of automated medical diagnosis based on the description of biomedical systems using two parameters: energy, reflecting the interaction of its elements, and entropy characterizing the organization of the system. The violations of the energy-entropy...

Full description

Bibliographic Details
Main Authors: E. A. Balashova, V. V. Bityukova
Format: Article
Language:Russian
Published: Voronezh state university of engineering technologies 2018-03-01
Series:Vestnik Voronežskogo Gosudarstvennogo Universiteta Inženernyh Tehnologij
Subjects:
Online Access:https://www.vestnik-vsuet.ru/vguit/article/view/1634
id doaj-41a295fd34bd48418b4ecd26ad8d779f
record_format Article
spelling doaj-41a295fd34bd48418b4ecd26ad8d779f2021-07-29T08:05:10ZrusVoronezh state university of engineering technologies Vestnik Voronežskogo Gosudarstvennogo Universiteta Inženernyh Tehnologij2226-910X2310-12022018-03-01794758110.20914/2310-1202-2017-4-75-811305Heuristic approach to the classification of postpartum endometritis and its formsE. A. Balashova0V. V. Bityukova1Voronezh state university of engineering technologiesVoronezh state medical institute named after BurdenkoТhe work is dedicated to the development of a method of automated medical diagnosis based on the description of biomedical systems using two parameters: energy, reflecting the interaction of its elements, and entropy characterizing the organization of the system. The violations of the energy-entropy cycle of biomedical systems is reflected in the symptoms of the disease. Statistical link between the symptoms of the condition of the body and the nature of excitation of its elements best expressed in the heuristic description of the system state. High accuracy classification of the patient's condition is achieved by using heuristic detection methods. In the proposed approach, allowing to estimate the probability of correct diagnosis increases the accuracy of the classification, and the estimated minimum amount of training samples and the capacity of its constituent signs. Classification technique consists in averaging the characteristic values in the selected classes, the preparation of the complex of symptoms of the most important signs of the disease, to conduct a "rough" diagnostic threshold rules that allow to distinguish severe forms of the disease, then differential diagnosis the severity of the disease. The proposed method was tested for classification of the forms of puerperal endometritis (mild, moderate, severe). The training sample contained 70 case histories. Syndrome to classify the patient's condition was composed of 17 characteristics. Threshold diagnosis has allowed to establish the presence of disease and to separate heavy. Differential diagnosis was used for classification of mild and moderate severity of postpartum endometritis. The accuracy of the classification of forms of postpartum endometritis amounted to 97.1%.https://www.vestnik-vsuet.ru/vguit/article/view/1634heuristic methods, classification status, postpartum endometritis
collection DOAJ
language Russian
format Article
sources DOAJ
author E. A. Balashova
V. V. Bityukova
spellingShingle E. A. Balashova
V. V. Bityukova
Heuristic approach to the classification of postpartum endometritis and its forms
Vestnik Voronežskogo Gosudarstvennogo Universiteta Inženernyh Tehnologij
heuristic methods, classification status, postpartum endometritis
author_facet E. A. Balashova
V. V. Bityukova
author_sort E. A. Balashova
title Heuristic approach to the classification of postpartum endometritis and its forms
title_short Heuristic approach to the classification of postpartum endometritis and its forms
title_full Heuristic approach to the classification of postpartum endometritis and its forms
title_fullStr Heuristic approach to the classification of postpartum endometritis and its forms
title_full_unstemmed Heuristic approach to the classification of postpartum endometritis and its forms
title_sort heuristic approach to the classification of postpartum endometritis and its forms
publisher Voronezh state university of engineering technologies
series Vestnik Voronežskogo Gosudarstvennogo Universiteta Inženernyh Tehnologij
issn 2226-910X
2310-1202
publishDate 2018-03-01
description Тhe work is dedicated to the development of a method of automated medical diagnosis based on the description of biomedical systems using two parameters: energy, reflecting the interaction of its elements, and entropy characterizing the organization of the system. The violations of the energy-entropy cycle of biomedical systems is reflected in the symptoms of the disease. Statistical link between the symptoms of the condition of the body and the nature of excitation of its elements best expressed in the heuristic description of the system state. High accuracy classification of the patient's condition is achieved by using heuristic detection methods. In the proposed approach, allowing to estimate the probability of correct diagnosis increases the accuracy of the classification, and the estimated minimum amount of training samples and the capacity of its constituent signs. Classification technique consists in averaging the characteristic values in the selected classes, the preparation of the complex of symptoms of the most important signs of the disease, to conduct a "rough" diagnostic threshold rules that allow to distinguish severe forms of the disease, then differential diagnosis the severity of the disease. The proposed method was tested for classification of the forms of puerperal endometritis (mild, moderate, severe). The training sample contained 70 case histories. Syndrome to classify the patient's condition was composed of 17 characteristics. Threshold diagnosis has allowed to establish the presence of disease and to separate heavy. Differential diagnosis was used for classification of mild and moderate severity of postpartum endometritis. The accuracy of the classification of forms of postpartum endometritis amounted to 97.1%.
topic heuristic methods, classification status, postpartum endometritis
url https://www.vestnik-vsuet.ru/vguit/article/view/1634
work_keys_str_mv AT eabalashova heuristicapproachtotheclassificationofpostpartumendometritisanditsforms
AT vvbityukova heuristicapproachtotheclassificationofpostpartumendometritisanditsforms
_version_ 1721258198068887552