DOI: 10.21045/2071-5021-2016-50-4-9
Narkevich A.N.1, Serov A.A.2, Vinogradov K.A.1, Narkevich A.A.1, Shadrin K.V.1
1Krasnoyarsk State Medical University named after
Professor V. F. Voyno-Yasenetsky, Ministry of Health of the Russian
Federation, Krasnoyarsk
2Tver State University, Ministry of Education and Science of the Russian Federation, Tver
Contacts: Artem N. Narkevich, e-mail:
This e-mail address is being protected from spam bots, you need JavaScript enabled to view it
Abstract. Significance. In the context of severe
shortage of budgets at all levels, planning expenditures on certain
areas of the national economy including healthcare, is extremely
relevant.
The purpose of the study was to explore capabilities of
self-organizing Kohonen maps for clustering regions according to their
morbidity rates with socially significant diseases to identify regions
that require strengthening of measures to control socially significant
diseases.
Methods. To cluster regions we used national statistics
on morbidity with socially significant diseases in 2006 and 2012 in 79
regions of the Russian Federation. Regions were clustered using Kohonen
self-organizing maps, implemented in BaseGroup Labs Deductor Studio
computer platform. According to the results the regions were divided
into 3 clusters: favorable, in transition and problem clusters.
Results. According to results of the cluster analysis
of regions in 2006, situation with socially significant diseases in
Russia is rather favorable; the majority of regions were assigned to the
transition cluster. Based on results of the 2012 clustering the
following conclusion was made: problem clusters and clusters in
transition tend to shrink, while the number of the regions in the
favorable cluster grows. Target boundary indicators were identified to
change cluster for a more favorable one as exemplified by the
Krasnoyarsk and Tver regions.
Conclusions. The considered approach towards clustering
regions according to data on morbidity with socially significant
diseases using Kohonen self-organizing maps helps to determine current
state of the region regarding these indicators, tendency towards
changing position of the region in the Russian Federation and to
identify target boundary indicators to improve situation with socially
significant diseases.
Scope of application. The proposed approach to cluster
regions and define boundary indicators can be used to identify regions
that require development of measures to control specific socially
significant diseases.
Êëþ÷åâûå ñëîâà: êëàñòåðèçàöèÿ ðåãèîíîâ; èñêóññòâåííûå íåéðîííûå ñåòè; ñîöèàëüíî-çíà÷èìûå çàáîëåâàíèÿ; ñàìîîðãàíèçóþùèåñÿ êàðòû Êîõîíåíà.
Keywords: regions’ clustering; artificial neural network; socially significant diseases; Kohonen self-organizing maps.
References
- Artyukhov I.P., Shul'min A.V., Denisov V.S., Kozlov V.V.
Planirovanie resursov zdravookhraneniya munitsipal'nogo obrazovaniya s
funktsioniruyushchim vakhtovym poselkom [Planning of health care
resources of the municipality with a functioning shift camps]. Sibirskoe meditsinskoe obozrenie 2012;(5):78-81. (In Russian)
- Botvin G., Zaboev M. Klasterizatsiya stran po makroekonomicheskim
pokazatelyam s ispol'zovaniem apparata iskusstvennykh neyronnykh setey
[Clustering country performance using artificial neural networks] // RISK: resursy, informatsiya, snabzhenie, konkurentsiya 2011;(4):552-556. (In Russian)
- Kulichenko V.P., Polubentseva E.I., Rakhaeva I.V., Chertukhina O.B.
Planirovanie okazaniya meditsinskoy pomoshchi, kak instrument
upravleniya sistemoy zdravookhraneniya regiona [Planning for health
care, as a management tool of the health system in the region] // Vestnik Sankt-Peterburgskogo universiteta. Seriya 11: Meditsina 2011;(1):190-200. (In Russian)
- Moskvicheva M.G., Shchepilina E.S., Shchetinin V.B., Yakushev A.M.,
Savishcheva I.P. Analiz sostoyaniya zdorov'ya naseleniya kak osnova
planirovaniya meditsinskoy pomoshchi na regional'nom urovne [Analysis of
the health status of the population as a basis for health care planning
at the regional level] // Rossiyskaya akademiya meditsinskikh nauk.
Byulleten' natsional'nogo nauchno-issledovatel'skogo instituta
obshchestvennogo zdorov'ya 2014;(2):86-91. (In Russian)
- Ob utverzhdenii perechnya sotsial'no znachimykh zabolevaniy i
perechnya zabolevaniy, predstavlyayushchikh opasnost' dlya
okruzhayushchikh [Approval of the list of socially significant diseases
and the list of diseases that pose a danger to others] postanovlenie
Pravitel'stva Rossiyskoy Federatsii ot 01.12.2004 g. ¹715. [Online].
2004 [cited 2015 Feb 14]. Available from:
http://base.garant.ru/12137881/ (In Russian).
- Seregin V.I. Opredelenie prioritetov v voprosakh meditsinskoy
profilaktiki neinfektsionnykh zabolevaniy na regional'nom urovne (po
materialam oprosa v sisteme pervichnoy mediko-sanitarnoy pomoshchi)
[Prioritizing issues of medical prevention of noncommunicable diseases
at the regional level (based on the survey in primary care)]. Sotsial'nye aspekty zdorov'ya naseleniya
[serial online] 2014 [cited 2015 Feb 14]; 37(3). Available from:
http://vestnik.mednet.ru/content/view/571/30/lang,ru/ (In Russian).
- Khaykin S. Neyronnye seti. Polnyy kurs [Neural Networks. Full course]. – M.: Vil'yams, 2006. – 1104s. (In Russian)
- Park S.H., Hosoishi S., Ogata K., Kuboki Y. Clustering of ant
communities and indicator species analysis using self-organizing maps. Comptes rendus biologies 2014;337(9):545-52.
- Taşdemir K., Milenov P., Tapsall B. Topology-based hierarchical clustering of self-organizing maps. IEEE transactions on neural networks 2011;22(3):474-85.
Views: 12210
|