Climatology
Ali Mohammad khorshiddoust; Kaveh Mohammadpour; Seyed Asaad Hosseini
Abstract
Introduction Prediction of hospital admissions related to climatic parameters is discussed matters that in recent decades in result from climate change, urbanization and air pollution has triggered widespread in many societies. Fluctuations in climatic parameters, in turn, can have a significant impact ...
Read More
Introduction Prediction of hospital admissions related to climatic parameters is discussed matters that in recent decades in result from climate change, urbanization and air pollution has triggered widespread in many societies. Fluctuations in climatic parameters, in turn, can have a significant impact on mortality and mortality, and the use of predictive models can be used to identify fluctuations in climatic parameters affecting disease and their prevalence and planning and Compatibility with the environment to be effective. Methodology Using of predictive models can be consider as an effective tool in managing and controlling the diseases, reducing mortality and planning. Recent study used from Artificial Neural Networks and Logistic Regression models as an effective tool in the prediction of nonlinear processes to predict the rate of asthma admissions related to Climatic parameters in Sanandaj/Sine city. Used data during period of 8-years (2001-2008) collected from synoptic station and Toheid and Beasat hospitals in the Sanandaj/Sine city. Then, the climatic parameters and rate of asthma admissions considered as an input and output data of models, respectively. Result and Discussion The results of the output of two nonlinear models of artificial neural network and Logit in examining the effect of climatic parameters on the number of the asthma patients in Sanandaj/Sine showed that the monthly average parameters with high coefficient of determination (R2=0.98) of temperature (average, minimum, maximum) and QFE pressure in the artificial neural network model and The monthly average minimum temperature, QFF pressure and wind speed (in Knot) in the Logit model have had the greatest impact on the rate of asthma admissions in the city. As the wind speed in the Logit model is more effective than other climatic parameters, that it is clear with the logarithmic superiority (-0.977) and the Wald coefficient (85.616). In general, air pressure, temperature and wind speed are the most effective climatic parameters on the number of asthma patients visiting the hospital. Therefore, depending on the accuracy of the models, the above argument means that among the parameters examined, the elements are more important than other parameters in the city. As the climatic elements have a more effective role in the admission patients to the hospital, and their fluctuations will be more significant in patients' fluctuations. The effects of environmental parameters (climate and pollutants) on diseases have previously been investigated as well, so that the results of previous logistic regression have display a increase respiratory disease, vulnerability of children to asthma and an increase in allergies; In the present study, the results of Logit model (69.5%) also indicate that decrease in the average minimum temperature lead to a decrease in the number of the asthma patients, it means that the rate of asthma is more less in temperatures close to zero or higher and vice versa, the admission more higher in the colder temperature (below zero); in the other words, the more balanced the temperature has the lower the rate, and in the colder the ambient temperature has the highest the number of asthma patients. Thus, comparison the present results and previous studies show that admissions change depending on climate, geographic position and the fluctuation of the elements and then the specific geographical location and the different climatic types of a region will play a decisive role in the number of asthma visitors to hospital. Conclusion The results indicated that Artificial Neural Network model predicted the asthma admissions related to monthly minimum, maximum and average temperatures with considerable accuracy, so that the correlation between actual and predicted data is significant with 0.01 coefficient and 0.99 confidence. Also, Input parameters in the Logit method shows that the rate of asthma admissions affected by parameters of average minimum temperature, average pressure QFF and average wind speed (in knot). In other words, the logarithmic ratio of each of cited parameters is significant with β-coefficients (-0.517), (-0.734) and (-0.977), respectively, that throughout of studied parameters is wind element of effective in asthma admissions then others to the hospital. In general, Artificial Neural Network model showed more sufficiency and accuracy than Logit model. As a result, both Logistic Regression and the Artificial Neural Network methods show that climatic parameters have a greater than 50% effect on the number of asthma patients referred to the hospital (the accuracy models: 69.5 and 98, respectively). In the Artificial Neural Network model, the most accurate possible result shows the more effective role of climatic parameters of temperature and air pressure on the asthma patients. Also, filtering the parameters examined at the output of the Logistic model showed the most possible coefficients for minimum temperature, QFF air pressure and wind speed (knot), among which wind speed was the most important element. Finally, the accuracy of the models showed that the Artificial Neural Network model has a higher accuracy depending on the coefficient of determination and highest correlation. Thus, Artificial Neural Network and Logit as nonlinear methods could well predict the relationship between climatic parameters and the number of the asthma patients. Also, according to the appropriate selection of input parameters and determination of different structures in the neural network is possible to design different models with the highest efficiency and can be considered as an effective and powerful tool in estimating similar studies. Introduction Prediction of hospital admissions related to climatic parameters is discussed matters that in recent decades in result from climate change, urbanization and air pollution has triggered widespread in many societies. Fluctuations in climatic parameters, in turn, can have a significant impact on mortality and mortality, and the use of predictive models can be used to identify fluctuations in climatic parameters affecting disease and their prevalence and planning and Compatibility with the environment to be effective. Methodology Using of predictive models can be consider as an effective tool in managing and controlling the diseases, reducing mortality and planning. Recent study used from Artificial Neural Networks and Logistic Regression modelsasan effective toolinthe prediction ofnonlinearprocessesto predict the rate of asthma admissions related to Climatic parameters in Sanandaj/Sine city. Used data during period of 8-years (2001-2008) collected from synoptic station and Toheid and Beasat hospitals in the Sanandaj/Sine city. Then, the climatic parameters and rate of asthma admissions considered as an input and output data of models, respectively. Result and Discussion The results of the output of two nonlinear models of artificial neural network and Logit in examining the effect of climatic parameters on the number of the asthma patients in Sanandaj/Sine showed that the monthly average parameters with high coefficient of determination (R2=0.98) of temperature (average, minimum, maximum) and QFE pressure in the artificial neural network model and The monthly average minimum temperature, QFF pressure and wind speed (in Knot) in the Logit model have had the greatest impact on the rate of asthma admissions in the city. As the wind speed in the Logit model is more effective than other climatic parameters, that it is clear with the logarithmic superiority (-0.977) and the Wald coefficient (85.616). In general, air pressure, temperature and wind speed are the most effective climatic parameters on the number of asthma patients visiting the hospital. Therefore, depending on the accuracy of the models, the above argument means that among the parameters examined, the elements are more important than other parameters in the city. As the climatic elements have a more effective role in the admission patients to the hospital, and their fluctuations will be more significant in patients' fluctuations. The effects of environmental parameters (climate and pollutants) on diseases have previously been investigated as well, so that the results of previous logistic regression have display a increase respiratory disease, vulnerability of children to asthma and an increase in allergies; In the present study, the results of Logit model (69.5%) also indicate that decrease in the average minimum temperature lead to a decrease in the number of the asthma patients, it means that the rate of asthma is more less in temperatures close to zero or higher and vice versa, the admission more higher in the colder temperature (below zero); in the other words, the more balanced the temperature has the lower the rate, and in the colder the ambient temperature has the highest the number of asthma patients. Thus, comparison the present results and previous studies show that admissions change depending on climate, geographic position and the fluctuationof the elements and then the specific geographical location and the different climatic types of a region will play a decisive role in the number of asthma visitors to hospital. Conclusion The results indicated that Artificial Neural Network model predicted the asthma admissions related to monthly minimum, maximum and average temperatures with considerable accuracy, so that the correlation between actual and predicted data is significant with 0.01coefficient and0.99 confidence.Also, Input parameters in the Logit method shows that the rate of asthma admissions affected by parameters of average minimum temperature, average pressure QFF and average wind speed (in knot). In other words, the logarithmicratio ofeach of citedparametersissignificant with β-coefficients (-0.517), (-0.734)and(-0.977), respectively, thatthroughoutofstudied parametersis windelement of effective in asthma admissionsthen others to thehospital. In general, ArtificialNeural Networkmodelshowed more sufficiencyandaccuracy than Logitmodel. As a result, both Logistic Regression and the Artificial Neural Network methods show that climatic parameters have a greater than 50% effect on the number of asthma patients referred to the hospital (the accuracy models: 69.5 and 98, respectively). In the Artificial Neural Network model, the most accurate possible result shows the more effective role of climatic parameters of temperature and air pressure on the asthma patients. Also, filtering the parameters examined at the output of the Logistic model showed the most possible coefficients for minimum temperature, QFF air pressure and wind speed (knot), among which wind speed was the most important element. Finally, the accuracy of the models showed that the Artificial Neural Network model has a higher accuracy depending on the coefficient of determination and highest correlation. Thus, Artificial Neural Network and Logit as nonlinear methods could well predict the relationship between climatic parameters and the number of the asthma patients. Also, according to the appropriate selection of input parameters and determination of different structures in the neural network is possible to design different models with the highest efficiency and can be considered as an effective and powerful tool in estimating similar studies.
