ISSN 0021-3454 (print version)
ISSN 2500-0381 (online version)

vol 64 / September, 2021

DOI 10.17586/0021-3454-2018-61-10-922-929

UDC 004.942


A. A. Semakova
ITMO University, Department of HighPerformance Computing;

N. E. Zvartau
Almazov National Medical Research Centre, Organizational Department on Cardiology and Cardiovascular Surgery; Head of the Department

S. . Kovalchuk
ITMO University; Junior Researcher

A. V. Boukhanovsky
ITMO University, Saint Petersburg, 197101, Russian Federation; Director

Read the full article 

Abstract. Arterial hypertension (AH) manifestation and development is associated with many endogenous and exogenous factors. A dynamic population model of AH development in different time scales with consideration of personalized treatment strategies is presented. The overall dynamics of the hypertensive patient group using a demographic model is created. Simulation of AH development at the individual level, is performed with the use of clustering the patients array based on data stored in electronic health records. For patients from each group, models of variability of the quantitative characteristics of the patient's digital profile were developed and the probabilities of the development of concomitant pathologies were estimated. A model is constructed of antihypertensive treatment tactics and strategy using a classifier predicting the most effective class of drugs for the patient based on his/her digital profile individual components collected before the treatment initiation. The developed simulation technology can be used to optimize the health care system processes, primarily for outpatient care, to estimate the effectiveness of new drugs and treatment strategies introduction, and to simulate scenarios for new clinical guidelines introduction. 
Keywords: arterial hypertension, dynamic population model, digital patient profile, clustering, CART-algorithm

  1. Polak S. and Mendyk A. Appl. Soft Comput., 2008, no. 1(8), pp. 734–739.
  2. Teixeira P.L. et al. J. Am. Med. Informatics Assoc., 2017, no. 1(24), pp. 162–171.
  3. Echouffo-Tcheugui J.B., Batty G.D., Kivimä Ki M., and Kengne A.P. PLoS ONE, 2013, no. 7(8),  pp. e67370.
  4. Farran B., Channanath A.M., Behbehani K., and Thanaraj T.A. BMJ Open, 2013, no. 5(3),  pp. e002457.
  5. Khalilia M., Chakraborty S., and Popescu M. BMC Med. Inform. Decis. Mak., 2011, no. 1(11), pp. 51.
  6. Mancia G. et al. Eur. Heart J., 2013, no. 28(34), pp. 2159–2219.
  7. Friedman C. Med. Informatics Knowl. Manag. Data Min. Biomed., 2005, pp. 423–448.
  8. Boykov I.V., Loginov A.Yu. University proceedings. Volga region. Physical and mathematical sciences, 2008, no. 2, pp. 15–20. (in Russ.)
  9. Lewington S., Clarke R., Qizilbash N., Peto R., and Collins R. The Lancet, 2002, no. 9349(360),  pp. 1903–1913.
  10. Van Der Maaten L. and Hinton G. J. Mach. Learn. Res., 2008, no. 9, pp. 2579–2605.
  11. de Amorim R.C. and Hennig C. Inf. Sci. (Ny)., 2015, no. 324, pp. 126–145.
  12. Semakova A., Zvartau N., Bochenina K., and Konradi A. Procedia Comput. Sci., 2017, no. 121,  pp. 852–858.