Neural Network of the gestosis diagnosis system

Authors

DOI:

https://doi.org/10.20535/RADAP.2018.72.42-46

Keywords:

pregnancy, cardio-vascular system, preeclampsia, the hemodynamic parameters, neural network

Abstract

Introduction. The work is devoted to the increase of the information content of the methods of express diagnostics of the state of the cardiovascular system of pregnant women by developing a diagnostic monitoring system and assessment of the state of hemodynamics of pregnant women due to the use of neural network technologies.
Currently, gestosis is one of the most urgent problems of modern obstetrics due to the prevalence and complexity of etiopathogenesis, the absence of early and reliable diagnostic criteria, effective prevention and treatment measures, high maternal and perinatal morbidity and mortality, as well as large economic costs of intensive care for patients. The proposed approach to optimization of the diagnosis of preeclampsia using registration key hemodynamic parameters, which allows to objectively evaluate the hemodynamics in pregnant women with preeclampsia, to determine the type of hemodynamics in pregnant women and to monitor the effectiveness of the therapy.
Neural network technologies for the analysis and processing of medical data. The synthetic structure of the artificial neural network has shown the effectiveness of its use for diagnosis of gestation, using real clinical data. The basis of the diagnostic system was the artificial neural network, using the application package Statistica.
Hybrid network as an adaptive system of neuro-fuzzy output system. In order to diagnose gestosis, we synthesized a neural network that allowed us to classify the pathologies of pregnant women on the basis of oscillometric data, using neural-fuzzy modeling in the Matlab environment. In the framework of this work, the ANFIS editor was used to build the model that would help create or load a specific model of the adaptive neuro-fuzzy inference system, perform its training, visualize its structure, change and adjust its parameters, and use the configured network to obtain the results of fuzzy inference. The findings of the tests, show that with the introduction of 7 basic parameters characterizing the state of a pregnant woman, the produced result is in the form of a definition of the type of hemodynamics and the degree of gestosis. The determination of the type of hemodynamics and the degree of gestosis allows the diagnostician and the obstetrician-gynecologist to assess the critical state of the pregnant woman at the time of the study and to take the necessary treatment measures.
Conclusion. In order to diagnose gestosis, a neural network was synthesized, which allowed classifying the pathology of pregnant women using neuro-fuzzy simulation in a Matlab environment. The application of the proposed system helps choose the tactics of treating a patient diagnosed with gestosis according to individually selected therapy, monitoring its effectiveness, which will have a positive effect on the course and outcome of pregnancy; most importantly, controlling the state of maternal hemodynamics will eliminate the unreasonable use of medications.

Author Biographies

M. P. Mustetsov, V. N. Karazin Kharkiv National University


Mustetsov N. P.

S. O. Bahan, Kharkiv National University of Radio Electronics

Bahan S. A.

References

Перелік посилань

Медведь В.И. Основные вопросы экстрагенитальной патологии / В.И. Медведь // Медицинские аспекты здоровья женщины. - 2011. - № 6. - с. 5-11.

Карпов А.Ю. Экспертная скрининг система: Экспресс-оценка системы кровообращения у беременных / А.Ю. Карпов, М.Б. Охапкин, В.И. Шмелев // Всероссийский форум: Интеллектуальные ресурсы регионов России на рубеже тысячелетий. - Ярославль. - 2000. - с. 70-72.

Киселева Н. И. Актуальные проблемы гестоза (патогенез, диагностика, профилактика и лечение) / Н. И. Киселева, С. Н. Занько, А. П. Солодков. - Витебск : ВГМУ, 2007. - 196 с.

Охапкин М. Б. Преэклампсия: гемодинамический адаптационный синдром / М.Б. Охапкин , В.Н. Серов, В.О. Лопухин // АГ-инфо. - 2002. - № 3. - С. 9-12.

Глухова Т. Н. Патогенетическое обоснование принципов диагностики, прогнозирования и комплексной терапии гестоза / Т.Н. Глухова, И.А. Салов, Н.П. Чеснокова. - Саратов : СарГМУ, 2005. - 47 с.

Халафян А. А. STATISTICA 6. Статистический анализ данных / А.А. Халафян. - М. : Бином-Пресс, 2007. - 512~с.

Боровиков В. П. Нейронные сети. Statistica Neural Networks. Методология и технологии современного анализа данных / В.П. Боровиков. - М. : Горячая линия - Телеком, 2008. - 392 с.

Mustetsov N. P. The possibilities of neural network technologies in solving medical problems // N.P. Mustetsov, S.A. Bahan / European Conference on Innovations in Technical and Natural Sciences, Vienna, Austria, 20 July 2017. - pp. 111-116.

References

Medved' V.I. (2011) Osnovnye voprosy ekstragenital'noi patologii [The main issues of extragenital pathology]. Medychni aspekty zdorov'ia zhinky, No 6, pp. 5-11.

Karpov A. U., Shmelev V. I. and Okhapkin M. B. (2000) Ekspertnaya skrining sistema: Ekspress-otsenka sistemy krovoobrashcheniya u beremennykh [Expert screening system: Rapid assessment of the circulatory system in pregnant women]. Materialy II Rossiiskogo foruma: Mat' i ditya, pp.70-72.

Kiseleva N. I., Zan'ko S. N. and Solodkov A. P. (2007) Aktual'nye problemy gestoza (patogenez, diagnostika, profilaktika i lechenie) [Actual problems of gestosis (pathogenesis, diagnosis, prevention and treatment)]. Vitebsk, VGMU, 196 p.

Okhapkin M. B., Serov V.N. and Lopukhin V.O. (2002) Preeklampsiya: gemodinamicheskii adaptatsionnyi sindrom [Preeclampsia: hemodynamic adaptation syndrome]. AG-info, No 3, pp. 9-12.

Glukhova T.N., Salov I. A. and Chesnokova N. P. (2005) Patogeneticheskoe obosnovanie printsipov diagnostiki, prognozirovaniya i kompleksnoi terapii gestoza [Pathogenetic substantiation of the principles of diagnosis, prognosis and complex therapy of gestosis], Saratov, SarGMU, 47 p.

Khalafyan A.A. (2007) STATISTICA 6. Statisticheskii analiz dannykh [STATISTICA 6. Statistical data analysis], Moscow, Binom-Press, 512 p.

Borovikov V.P. (2008) Neironnye seti. Statistica Neural Networks. Metodologiya i tekhnologii sovremennogo analiza dannykh [STATISTICA Neural Networks: Methodology and technology of modern data analysis]. Moscow, Goryachaya liniya, 392 p.

Mustetsov N.P. and Bahan S.A. (2017) The possibilities of neural network technologies in solving medical problems. European Conference on Innovations in Technical and Natural Sciences, p.111-116.

Published

2018-03-30

How to Cite

Мустецов, М. П. and Баган, С. О. (2018) “Neural Network of the gestosis diagnosis system”, Visnyk NTUU KPI Seriia - Radiotekhnika Radioaparatobuduvannia, 0(72), pp. 42-46. doi: 10.20535/RADAP.2018.72.42-46.

Issue

Section

Radioelectronics Medical Technologies