CC BY
0MEDICAL SCIENCES
DEVELOPMENT OF SYSTEM THAT ANALYZES AND FORECASTS THE EPIDEMIOLOGY OF
MYCOSES
Kerimov S.G.
Azerbaijan State Advanced Training Institute for Doctors named after A. Aliyev
Azerbaijan, Baku city Abdullayeva G. G.
Institute of Control Systems of Azerbaijan National Akademy of Sciences
Azerbaijan, Baku city Mirzoyev A.H.
Azerbaijan State Advanced Training Institute for Doctors named after A. Aliyev
Azerbaijan, Baku city
Abstract
The epidemiology of mycoses was studied in different regions of the Republic of Azerbaijan. This problem was confirmed by studies conducted in various organizations in the cities of Baku, Ganja, Lankaran and Kurdamir, the current situation and future forecast was developed in the STATISTICA-7 software package through a time series model.
Keywords: mathematical model, mycoses, random variables, time series, trend line, biostatistics methods
1. Introduction. As a model object, the source object should be replaced in the process of research so that its direct investigation could give new knowledge about the source object. Modeling means the process of creation, study and application of the model. It is closely related to such categories as abstraction, analogy, assumption, etc. Process of modeling involves the construction of abstractions, judgments based on analogy and the creation of scientific hypotheses. Process of modeling consists of three elements: the subject (researcher), the object of study and the mediating model between the relationship of the cognizing subject and the cognizable object [1]. Obviously, the model loses its meaning when it is identical to the original (when it is no longer the original), and also when it has completely different characteristics. A characteristic feature of the model is the possibility of carrying out various experiments on it, which corresponds to ethical and social norms. In the field of medicine, such models lead to a large number of imitations, experiments cost bloodless and cheaper. Models have been developed for many medical illnesses. For example, A.N.Bakin and S.Y.Khripkov [2] developed a mathematical model based on probability theory of spread of epidemiological processes associated with several infectious diseases (hepatitis A, tularemia, botulism and anthrax) and studied the dynamics of these diseases. O.Y. Karyakina and others [3] explained the advantages of mathematical models for several areas of practical medicine, covered with their help the summary information on solutions to strategic and tactical problems in the process of forecasting, differential diagnosis and treatment. It is noteworthy that there are no applied models for diagnosis and, especially, prediction of fungal diseases.
Here we note that the statistical methodology is actively used in medicine for more than a hundred years. Analyzing in his article "How it is impossible to collect and process statistical information in medicine in order to avoid receiving erroneous results, instead of
real ones" the possibility of applying mathematics in medicine, M.K. Zenes came to the conclusion that "the methods of mathematical statistics will occupy a worthy place in medicine regarding the management of human health, because it is possible and necessary to wait for the most productive results from their application".
2. Statement and substantiation of issue. 20% of the world's population have superficial fungal skin diseases. A study in 16 European countries in 2003 showed that 35% of the 70 000 population are infected with fungal skin diseases. In general, 2.5 million people suffer from fungal diseases, 30-40% of which falls on the share of superficial fungal skin diseases. Over the past 10 years, the amount of superficial fungal skin diseases has increased by 2.5% and every year this increase is 5%. Among superficial fungal skin diseases, dermisophytes dominate and are found in 10% of the world's population. (Schmid-Wendtner M.H. - 2008; Sergeyev A.Y, Ivanov O.A, Sergeyev Y.B. - 2005, Frizin V.V. and etc. -2009; Pek Y.Y., Korysheva V.G., Chipina G.A., Ignatieva S.M. - 2010)
The fact that dermatomycosis as a fungal skin disease is one of the most common skin diseases, the organization of prevention and treatment of these diseases is at the level of state priorities. Dermatovenerologic dispensaries, dermatovenerologic rooms in the city and district hospitals, hygiene and epidemiology centers in Azerbaijan held various types of activities in this area. However, these complexes of measures did not cause a decrease in fungal diseases of the skin. Conducted statistical studies showed that in the regions of Azerbaijan and Baku in recent years the incidence of fungal skin lesions has increased. Given the current climate, conditions, socio-economic situation in Azerbaijan and other factors, there is a need to improve more accurate, quality, effective methods of examination, diagnosis and treatment. Therefore, it is expedient to study the epidemiology of the most common superficial fungal skin diseases in Azerbaijan (in
different cities and regions), assess the current situation and make a forecast of their development. The article suggests an approach to solving this problem.
3. Study material. In connection with the problem
under consideration, the article selected four major
cities of Azerbaijan: Baku, Ganja, Lankaran and Kurdamir. Let's look at a few enterprises in these cities as the master sample.
In general, the pathological fungi can be classified as follows:
Fig.1. Classification ofpathological fungi
To conduct the experiment, Ganja Textile Factory, Ganja Fatoglu mill, Karamel confectionery and Firdevs confectionery were chosen as the selection sample from the sample of Gyandzhda city, and the Kurdmir milk processing plant and the Kurdamir Grain Processing Plant were selected from the Kurdamir sample.
The article considers such superficial skin diseases as T - trichophyte, C - candida, E - epidermyphyton, M - microsporia. To conduct an organized survey, a questionnaire and a patient card were drawn up in accordance with the requirements listed below: full name; date of birth; place of residence; living conditions; profession; working conditions; climatic conditions: a. temperature mode, b. humidity; health status: a. anamnesis morbi; b. anamnesis vitae; c. status lokalis; accompanying illnesses; microscopy; seeding, etc.
4. Study method. Process of modeling involves the construction of abstractions, judgments based on analogy and the creation of scientific hypotheses. It is obvious that the model loses its meaning when it is identical to the original (when it is no longer the original), and also when it has completely different characteristics. In medicine, there are three types of mathematical models: mathematical models of qualitative, imitative and quantitative type. Models of qualitative type are mainly used in scientific research.
Various experiments on the model are carried out by means of medals of imitation type. These models correspond to both educational and ethical and social norms. An example of this can serve as an individualized treatment model for snake poisoning [5].
5. Problem solution. To solve this problem, the article proposes to use parametric and nonparametric methods of time series and biostatistics [6].
The time series consists of two types of data:
> continuous data - allows to obtain the values of variables observed in unspecified time period;
> discrete data - allows to obtain the values of variables observed in specified time period.
The time series method is considered to be a suitable method to verify the dependence of human health on environmental pollution used in epidemiological studies, khortnye studies, that is, when conducting observations at certain equal intervals of time and in other studies [3]. This method can be used both individually and at the group level. Summing up, it can be said that the time series method can be used in any studies where observations and fixing of results are performed in equal intervals of time. For example, in the Table 1 are considered five large enterprises of the cities of Ganja and Kurdamir and monitoring of workers of these enterprises was carried out from spring 2013 to spring 2016.
Table 1
Fragment of the base, reflecting the existing sanitary and epidemiological condition of superficial fungal skin diseases in the cities of Ganja and Kurdamir in the period from autumn 2013 to spring 2016
Period, in seasons
№ Full name Date of birth Place of work Position 13- autumn r et e 'I 10 14- spring r e s s 3 m 10 c S 3 3 10 15- winter 15- spring 15- summer 15- autumn r et c 'g 10 16- spring
0 2 2 0 2 2 2 0 2 0 2 0 2 0 2 2 0 2
1 Hasanov K. 1989 Worker T T T T T T T
2 Ibrahimov G. 1991 Worker T T T T T T T T
3 Ismayilov T. 1992 Worker T T T T
4 Mamadova G. 1988 Weaver E E E E E E
5 Suleymanov N. 1979 laboratory technician T T T T T T
6 Kazimov I. 1979 Worker K K K K K
7 Guliyev T. 1987 Mechanic T T T T T
8 Verdiyev A. 1980 Mechanic
9 Aliyev I. 1978 Driver T T T T T
10 Aliyev E. 1980 Washman C C C C C C
11 Gojaliyev I. 1972 Electrician
12 Ibrahimov I. 1959 Welder T T T T T T
13 Khudiyev T. 1964 Packer T T T T T
14 Hasanov Z. 1976 Loader
15 Aliyev Sh. 1979 Driver T T T T T T T
16 Pashayev A. 1990 Driver T T T T
17 Isgandarov A. 1979 Worker T T T T T T T
18 Mammadov Y. 1969 Worker T T T
19 Orujov O. 1965 Equipment operator
20 Asgarov V. 1949 Machinist T T T T T
21 Garibov Z. 1959 Machinist
22 Aliyev R. 1970 Worker M
23 Ahmadova F. 1978 Ribbon tying E E E E
24 Volkan Kaya 1982 Miller T T T T T
Mustafayeva S. 1969 Winder C C C C C
This type of tables were collected and systematized during each of the four seasons of 20132016. The following signs were adopted in the table: T - trichophyte, C - Candida, E - epidermyphyton, M -microsporia.
587 people participated in the experiments, through random selection, pathological material was taken from 126 people from Ganja and 81 people from Kurdamir and a laboratory examination was carried out. Based on this information, a database was developed. In the result of frequency analysis was calculated the frequency of occurrence Nm/Num of various mycoses. Since according to the materials obtained from Kurdamir Nm/Num = 0,2389, and from Ganja Nm/Num = 0, 3156. The time series consists of 11 vectors. The analysis of the series and the prediction of the situation on this basis were realized in two versions: they were studied separately for each season and together for all seasons. Parametric and nonparamomatic statistical criteria were used to estimate changes between indicators within a time interval. But regarding the application of parametric criteria, it is advisable to use nonparametric methods because the refinement of the distribution of the indicator and the frequency of measuring the values of the indicators in a certain time interval can not be predetermined in advance. In contrast to the analysis of random samples, time series are based on observing data at equal intervals [4]. The analysis of time series has two purposes: to determine the nature of the series
and to predict it. In both cases, a series model was developed to interpret the data in question. Most regular time series variables are part of two classes: they are either trend or seasonal components. The dynamics of change reflects its trend. The trend is formed from a common systematic linear or nonlinear component that regularly changes with time. The seasonal component is repeated periodically. The use of time series for the forecast is due to the fact that the influence of certain factors on the data of the observed process in the past, in the present is similar to the same effect in the near future.
In the simple case, if the independent value Xi expresses environmental factors, then Y is the result of the impact. In this case, it is possible to create a regression relationship between them. This can also be done in a time sequence. The general regression equation for multidimensional time series is expressed as follows [7]:
E(Y) = a0 + Y(aifi(Xi)) + s,
where, Y - result of the impact (quantitatively); Xi- environmental factors; i - the index of the environmental factor (current number), i=1,2,..., n; n -the number of factors; E h f- the connection and transforming functions of series; ao ,...,a— coefficients of trends (primary values); e - residual limit (method error).
The problem is to determine the coefficients ai in the equation. The function E is determined by the type
of distribution of the dependent variable, the functionf At the first stage, we define the normal
has a filter function to eliminate the effect of trends. distribution of diseases T - trichophyte, C - candida, E
Now we will consider the results of time series - epidermyphyton, M - microsporia. analysis using the software package Statistics-7.
Line Plot of Var3
Spreadsheet2 10v*18c Line Plot of Var2
01 23456789
— / \
/ \
/ \ / \
___ / \ /
— \ / \ / \
01 23456789
Fig.2. Curves of normal distribution
At the second stage, we determine the seasonal - epidermyphyton, M dependence of diseases T - trichophyte, C - candida, E monitoring period.
microsporia during the
Line Plot of Var4 ■eadsheet2 10v*18c
At the third stage, we plot the smoothing curve for becomes possible to remove the "teeth" effect from the the time series of diseases T - trichophyte, C - candida, edges. E - epidermyphyton, M - microsporia. In this way, it
2,5
Exp. smoothing: S0=2,000 T0=0,000 Lin.trend,no season; Alpha= ,100 Gamma=,100 VAR4
A
\
] ,\4 •1 -> -
VIf \ / \ w y-" V \ 1
4 6 8 10 12 14 16 18 20 — VAR4 (L) --- Smoothed Series (L) ....... Resids (R)
58 56 54
Exp. smoothing: S0=57,04 T0=-,083 Lin.trend,no season; Alpha= ,100 Gamma=, VAR2
-----m-----
— VAR2 (L) --- Smoothed Series (L)
_J_
18 20 Resids (R)
64
62
60
52
50
48
22 24
4
22 24
Fig.4. Curves of smoothing
Analysis of the data shows that the frequency of Finally, at the fourth stage, we can calculate the
occurrence of the M-microsporia disease is equal to 1, prognosis of these diseases by the already observed for this reason, its smoothing function is not required. 2013-2016 years.
Forecasts; Model:(1,0,0) Seasonal lag: 12 hput: VAR3 : ln(x) Start of origin: 1 End of origin: 13
.. . 1. 1 .... 1 ...
\ *
\ \ ■ ■
\ 1 f
\ j
y
0 2 4 6 8
10 12 14 16 18 20 22 24 ■ - Forecast ■ ■ ■ ± 90,0000%
64 r
3,2
62
3,1
60
3,0
58
2,9
56
2,8
54
2,7 2,6 52 2,5 50 2,4 48 -
Forecasts; Model:(1,0,0) Seasonal lag: 12 Input: VAR2
Start of origin: 1 End of origin: 13
............
A
/ \ / i
• \ / \ / I ! 1 ! / ! \ ! / ! 1 !
11 1 j 1 j
\ / V \ / A|/ . Iji.....
V 1/ V
8 10 12 14 16 18 20 ■ Observed - - Forecast ■ ■ ■ ± 90,0000%
-| 64 - 62 - 60
- 58
- 56
- 54
- 52
- 50
- 48 22 24 26
Forecasts; Model:(1,0,1) Seasonal lag: 12 Input: VAR4 : ln(x); x+0,000 Start of origin: 1 End of origin: 13
A r\ \
/ \ / \ \ \ n » L /
V v \ / . # * *
i -- ""r i""r~
r
4 6 8 10 12 14 16 18 20 22 24 26 — Observed - - Forecast - - ± 90,0000%
Fig.5. Prognosis oof mycosis spread
For the above reason, the microsporium forecast curve has not been compiled.
As a result of our research and experiments, we have created a system that monitors the epidemiology
of fungal infections and determines the current status and prognosis through time series. Fig. 6 reflects the overall architecture of this system.
24
6
26
,5
1.5
1.0
1,0
-0,5
-0,5
-1,0
-1,0
-1,5
-1,5
0 2
Introducti on
-
at
"S
■a
M
s c
•E jd
о
"S3 a
о t
s a
a
00
s Ö
J 'Ü
1 e r
еь О
Statistical processing block
Decision-making block
1
Seasonal processing
General processing
Forecast
Recommendations
Fig.4. The architecture of the system that analyzes and predicts epidemiology of mycoses
6. Conclusion.
The article deals with the essence of the mathematical model in medicine, conducting studies using this model, diagnosis of mycoses by mathematical modeling (time series), sampling were performedin the cities of Ganja and Kurdamir, seasonal measurements were made during the period from spring 2013 to spring 2016, a database was created, time series and regression equations were compiled, as well as the analysis and forecast of T - trichophyte, C -candida, E - epidermitophyton, M - microsporia diseases has been carried out using the software package Statistics-7.
References
1. A.A Samarsky, A.P. Mikhailov. Mathematical Modeling: Ideas. Methods. Examples. Moscow: Fizmat, 2005, -320 pages.
2. A.N.Bakin, S.Y.Khripkov Mathematical modeling of the dynamics of the risk of infectious disease.// Risk problems. T.6, 2009, №2, pages.24-31
3. O.Y. Karyakina and others. Application of Mathematical Models in Clinical Practice. // Human Ecology. 2012, 7. pages103-106
4. Abdullaeva G.G., Ali Shakhintash, Imranov F.B. Diagnostics system of the functional state of ho-meostasis in toxicology (on the example of the bite of a Levantine viper)//Journal of Computational and Applied Mathematics, Kiev, 2011, №3(106)
5. V.I. Yunkerov, Grigoryev S.G. Mathematico-statistical processing of medical research data. St. Petersburg, 2002, page 212.
6. V.M. Guryeva, Y.B. Kotov. Analysis of short segments of time series in medical problems // Institute of Applied Mathematics named after M.V. Keldish, Moscow, 2005. This study was supported by the Russian Foundation for Basic Research (project No. 04-0100434)
7. I.I. Yeliseyeva. General Theory of Statistics: Textbook for Universities / I.I. Yeliseyeva, M.M. Yuzbashev; edited by I.I. Yeliseyeva. - Moscow: Finance and Statistics, 2009 - page 656.
ИММУННОЕ ЗВЕНО РАЗВИТИЯ ГЕНЕРАЛИЗОВАННОГО ПАРОДОНТИТА
Дроник I.I.
ВДНЗУ "Буковинський державний медичнийутверситет" асистент кафедри стоматологИ дитячого в1ку
IMMUNE PART OF GENERALIZED PPERIODONTITIS DEVELOPMENT
Dronyk I.I
HSEE of Ukraine «Bukovinian State Medical University», Department of Childhood Dentistry, Assistant
