ANALYSIS OF MEDICAL HISTORY AND LABORATORY PARAMETERS OF PATIENTS WITH ENDOMETRIAL HYPERPLASTIC PROCESSES USING MULTI-PARAMETER NEURAL NETWORC CLUSTERING
DOI:
https://doi.org/10.11603/1811-2471.2016.v27.i3.6779Keywords:
endometrial hyperplasia, perimenopausal age, simple endometrial hyperplasia, complex endometrial hyperplasia, disease progression, clustering neural network.Abstract
SUMMARY. To improve the efficiency of diagnosis of endometrial hyperplastic processes in perimenopausal women was conducted a retrospective analysis of 52 cards inpatient women who were treated at gynecological departments of Ternopil region. According to the histopathological finding 1st group included 28 women with simple endometrial hyperplasia; 2nd group – 24 women with complex endometrial hyperplasia. Comparison group consisted of 12 somatically healthy women. The average age of patients of group 1 – (47.0±1.0) years, group 2 – (56.0±1.5) years. In order depth analysis parameters neural network clustering applied. Established that the risk factors for hyperplastic processes of endometrium in women of perimenopausal period include disadvantaged socio-economic factors: residence in rural areas, difficult working conditions, bad habits and burdened obstetric and gynecological history, 3 or more pregnancies, a large number (4 and more) obstetrical interventions, use of intrauterine contraception. In order to determine the value of the combined changes of various parameters for the prediction of disease progression implemented neural network clustering. It is found that the likelihood of the progression of endometrial hyperplasia increases with age, if it is combined with the number of leukocytes in peripheral blood.
References
Vykhlyaeva E. M. Hyperplastycheskye protsessy éndometryya / E. M. Vykhlyaeva, B. Y. Zheleznoe, V. N. Zaporozhan // Rukovodstvo po éndokrynnoy hynekolohyy. M. : MYA, 2002. – S. 603–710.
Volkova N. Y. Zdorovʹe zhenshchyn v perymenopauze. Teoryya y praktyka / N. Y. Volkova // Zhur. akush. y zhen. bolezney. 2004. – Vyp. 2, T. LII. – S. 54–59.
Kuznetsov Y. V. Hyperplastycheskye protsessy éndometryya / Y. V. Kuznetsov. – M. : Medytsyna, 2009. – 48 s.
Klymakterycheskyy peryod y menopauza // Ratsyonalʹnaya farmakoterapyya v akusherstve y hynekolohyy : rukovodstvo dlya prakt. vr. – M., 2005. – Hl. 48. – S. 593–603.
Kontseptsiya informatyzatsiyi okhorony zdorovʺya Ukrayiny / O. P. Mintser, YU. V. Voronenko, L. YU. Babintseva [ta in.] // Medychna informatyka ta inzheneriya. – 2012. – № 3. – S. 5–29.
Martsenyuk V. P. Neyromerezheve prohnozuvannya skladannya studentamy-medykamy litsenziynoho intehrovanoho ispytu "Krok 1" na osnovi rezulʹtativ potochnoyi uspishnosti ta semestrovoho kompleksnoho testovoho ispytu / V.P. Martsenyuk, A.V. Semenetsʹ, O.O. Stakhansʹka // Medychna informatyka ta inzheneriya. – 2010. – № 2. – C. 57–62.
Metabolic and vascular effect of progestins in postmenopausal women: implications for cardioprotection / G. M. C. Rosano, C. Vitale, A. Silvestri, M. Fini // Maturitas. – 2003. – Vol. 46 (Suppl. 1). – P. 17–29.
Shepherd A. J. Second-Order Methods for Neural Networks : Fast and Reliable Training Methods for MultiLayer Perceptrons / A. J. Shepherd. – London : Springer, 1997. – 342 p.
