Combined Predictive Analysis Of Two Reference Values ??of Chinese Healthy Adult Blood And Geographical Environment

"Blood" is a red opaque viscous liquid flowing through the vessels,composed of plasma and blood cells,which circulates continuously under the action of the heart.The blood contains a variety of nutrients needed by the human body,the nutritional organization,regulation of organ activity and the role of resistance to harmful substances.Blood storage of human health information,so in many diseases diagnosis and treatment,the need of blood tests.Plasma fibrinogen and plasma D-dimer are indicators of coagulation function,in the current medical research,the main concern lies in the body's own height,weight,age,gender and other subjective factors on the impact of the index value,but very few studies are aimed at the impact of the geographical environment on the two indexes of blood.So this article selects Chinese healthy adults'two blood reference value for research,aims to analyze how geographical factors affect medical reference value and improve analytical methods when explore the connection between medical reference values and geographical factors.By consulting the literature,collects 10701 of Chinese healthy adults' reference values of Fib and 20795 of Chinese healthy adults' reference values of D-dimer,from 138 cities in 31privinces.25 geographical factors were slected to examined the significance between blood reference value and geographical factors by spatial autocorrelation,correlation analysis and grey correlation analysis,through the text of significance,the main factors of constructing the model are extracted,which are principal component model,the artificial neural network model,the support vector machine model,the ridge regression model.Because the single model has its own advantages and disadvantages,in order to make the prediction results more accurate,we choose 4 kinds of combination forecast model,and evaluate the prediction results of the combined model,to choose the best model,and forecast data for spatial mapping.Through the research,we can have this following conclusions:?1?The reference value of the two indexes of adult blood has spatial correlation,which is not independent,but is influenced by geographical factors,there is a regional.?2?Base on the correlation analysis,8 geographical factors have significant correlations with the reference value of adults' Fib,they were latitude,annual sunshine duration,annual mean air temperature,and so on;9 geographical factors had significant correlations with the reference value of adults' D-dimer,they were latitude,altitude,annual sunshine duration,and so on.?3?The sum of squares of errors prediction model is chosen as the best model to predict the adults' Fib reference value:??? =0.107598Y₁+0.242796Y₂+0.341051 Y₃+0.308555Y₄The optimal weighting prediction model is chosen as the best model to predict the adults' D-dimer reference values:???=0.542Y₂+0.458Y₃?4?The spatial distribution trend of Fib is high in the southeast,low in the northwest,high in the middle and high in the coastal area.The spatial distribution trend of D-dimer is high in the northwest and low in the southeast.In this paper,we selects Chinese healthy adults as research objects,analyzes the relationships between blood reference value and geographical factors.Through a variety of modeling methods,the two indicators of blood reference values were predicted,and the model was combined to improve the prediction accuracy.The distribution map can clearly show the geographical distribution of the reference value,providing a scientific basis for developing various Fib and D-dimer reference values in various areas.