Font Size: a A A

Combination Forecast Analysis Of Impulse Oscillometry System Reference And Geographical Factors

Posted on:2015-05-23Degree:MasterType:Thesis
Country:ChinaCandidate:R Y XueFull Text:PDF
GTID:2284330434451488Subject:Regional Environmental Studies
Abstract/Summary:PDF Full Text Request
Impulse oscillation system (IOS) of lung function examination is an important index to check lung respiratory mechanics which based on forced oscillation principle. It has important means on airway obstructive diseases’diagnosis and evaluation. It has four main indexes, respectively is Response frequency (Fres),5Hz frequency oscillation frequency of the airway resistance (R5),5Hz oscillation frequency of breathing reactance (X5) and respiratory impedance (Zrs). Including height, weight, age, gender factors, there are many factors that can affect the diagnosis in medicine. However, the geographical environment also influences the reference values of IOS significantly. At present, only a few researches’aim is to explore the relationship between the IOS reference value and geographical environment. Meantime, lack of domestic IOS reference standards is also a serious problem. So this article selects children and adults’four IOS reference value for research to make up the lack of overlook the geographical factors when formulate the medical references. And also this article aims to analyze how geographical factors affect healthy people’s IOS reference value, explore the principle how geographic factors affect medical reference value and improve analytical methods when explore the connection between medical reference values and geographical factors.This article collects6945of5to9years old healthy children and40to65years old healthy adults’ IOS reference values (Fres, R5, X5, Zrs) from all over China and grouping them into boy, girl and adult group bases on different genders and ages. This article selects25geographical factors. They are longitude, latitude, altitude, annual sunshine duration, annual average temperature, annual average relative humidity, annual rainfall, annual temperature range, annual average wind speed, topsoil gravel content, topsoil sand fraction, topsoil silt fraction, topsoil clay fraction, topsoil reference bulk density, topsoil bulk density, topsoil organic carbon, topsoil pH, topsoil CEC (clay), topsoil CEC (soil), topsoil base saturation, topsoil TEB, topsoil calcium carbonate, topsoil gypsum, topsoil sodicity (ESP), topsoil salinity (ECe). On the basis of correlation analysis, derived the correlative geographical factors from the database, using ridge regression analysis to make prediction and make comparison analysis. For analyzing the IOS indicators and the geographical factors’ complicated relationship better, using the artificial neural network, support vector regression machine model predict adult data for further analysis. Finally by utilizing the method of combination forecast model to combine ridge regression analysis, artificial neural network, support vector regression machine model, and build up the combination forecast model. Meanwhile, using statistical analysis and interpolation precision mapped the geographical distribution in each group and search the influence mechanism.This article outputs twelve IOS ridge regression prediction equation. The result shows that the four IOS reference values have relationship with geographical factors. It means that geographic differences may affect the lung tissue structure and function; four combination forecast model was constructed, and its prediction result is of high accuracy and practical value; healthy Chinese’s IOS indexes geographical distribution plots have been output and show the regional differences of IOS indexes. The analysis results show:Geographical factors affect children’s Fres reference values:longitude (-0.605,-0.593), altitude (0.349,0.365), annual sunshine duration (-0.553,-0.526), annual average wind speed (-0.522,-0.493), topsoil bulk density (0.351,0.350), topsoil organic carbon (0.338,0.334), topsoil CEC (clay)(-0.510,-0.533);Geographical factors affect children’s R5reference values:longitude (-0.367,-0.348), altitude (0.374,0.368), annual sunshine duration (-0.382,-0.347), topsoil CEC (clay)(-0.461,-0.448);Geographical factors affect children’s X5reference values:longitude (0.367,0.419), altitude (-0.416,-0.447), topsoil clay fraction (-0.388,-0.389), topsoil CEC (clay)(0.467,0.469);Geographical factors affect children’s Zrs reference values:longitude (-0.365,-0.336), altitude (0.351,0.340), annual average wind speed (-0.392,-0.365), topsoil bulk density (0.407,0.404), topsoil organic carbon (0.357,0.352), topsoil pH (-0.431,-0.446), topsoil calcium carbonate (-0.432,-0.436), topsoil gypsum (-0.491,-0.483), topsoil sodicity (ESP)(-0.480,-0.473), topsoil salinity (ECe)(-0.61,-0.554); Geographical factors affect adults’ Fres reference values:longitude (-0.440), altitude (0.396), annual average temperature (-0.225), annual average relative humidity (-0.228), annual average wind speed (-0.318), topsoil clay fraction (0.243), topsoil CEC (soil)(0.255).Geographical factors affect adults’ R5reference values:longitude (-0.319), altitude (0.225), annual average relative humidity (-0.286), annual average wind speed (-0.360), topsoil sand fraction (0.297), topsoil silt fraction (-0.424), topsoil gravel content (-0.233);Geographical factors affect adults’ X5reference values:longitude (0.232), altitude (-0.279), annual sunshine duration (0.283), topsoil bulk density (0.249), topsoil organic carbon (-0.298), topsoil CEC (soil)(-0.288);Geographical factors affect adults’Zrs reference values:longitude (-0.398), altitude (0.363), annual average temperature (-0.306), annual average relative humidity (-0.329), annual rainfall (-0.242), annual temperature range (0.245), annual average wind speed (-0.347), topsoil silt fraction (-0.364).This paper constructed the IO S reference ridge regression equation of relation with geographical factors. These provide a certain reference for medical examination. This paper brings combination forecast model which rarely use in geography to medical reference value and geographical factors’ relationships research, and gained good results no matter in the output of standard deviation value or fitting with the truthful data. These prove that combination forecast model has a good application prospect in geography. Choosing the combination of linear and nonlinear models when built the model, these also gained nice effect. The study on impact factors confirm that it will shows some differences as a result of the affect by different geographic factors when people living in it. The conclusion is that medical reference value has significant relationship with the regional environment, climate environment and soil environment which human living in.
Keywords/Search Tags:Impulse oscillation system, Geographical factors, Ridge regressionanalysis, Combination forecast model, Statistical analysis, Geostatistical analysis
PDF Full Text Request
Related items