中国人肺活量正常参考值与地理因素的关系

【作者】 何进伟；

【导师】 葛淼；

【作者基本信息】 陕西师范大学 ， 环境科学， 2009， 硕士

【摘要】 肺活量正常参考值是肺功能检查的一项重要指标,可以用来判断限制性通气功能障碍程度。在临床上,对肺活量正常参考值的增加或是减少的解释多以病理学为主,生理学解释也多以身高、体重、年龄以及性别等为主,很少涉及人类生存环境地形、气候、地貌等因素对肺部器官功能的影响。而本论文的创新之处在于从地理环境方面出发,对不同性别各年龄段的中国人肺活量正常参考值与地理因素,进行了线性与非线性的定量化研究与对比。经过充分的检索发现,虽有多篇论文报道了肺活量正常参考值与海拔、温度、湿度、风速等地理因素的关系,但还只是定性的描述或用线性回归分析的方法进行简单的线性分析,像本论文这样对肺活量正常参考值与地理因素进行专题化、定量化、线性与非线性化的研究,从整个地理环境着手,系统的以中国这一大的区域来研究,还未见报道。为探讨不同性别各年龄段肺活量正常参考值与所选七种地理因素复杂的关系,本论文通过大量检索文献,以及向有关单位求购的方法,收集了全中国各地区不同年龄不同性别的肺活量正常参考值。根据国家测绘局数据中心提供的共享资料,国家气象局数据中心提供的共享资料及有关地理著作、词典和相关文献,收集了七项地理因素指标:海拔高度（X₁）、年日照时数（X₂）、年平均气温（X₃）、年平均相对湿度（X₄）、年降水量（X₅）、气温年较差（X₆）、年平均风速（X₇）。对医学指标与这些地理学指标进行了线性和非线性的研究,从而建立它们之间的多元回归模型、偏最小二乘回归模型、岭回归模型、非线性回归模型、BP人工神经网络模型。最终通过模型对比研究,寻找出肺活量正常参考值的最优预测模型:①预测男性儿童肺活量正常参考值运用人工神经网络模型,构建5层神经网络,训练次数为264次时,模型最优;②预测女性儿童肺活量正常参考值运用偏最小二乘回归模型,其模型为:Y=1321.21-0.043514X₁+0.010103X₂-0.25246X₃-0.75417X₄-0.013527X₅+3.83186X₆+38.61289X₇±133.50;③预测青春期男性肺活量正常参考值运用偏最小二乘回归模型,其模型为:（?）=3024.77-0.131640X₁+0.022388X₂+2.381548X₃-1.344117X₄-0.030850X₅+7.881706X₆+39.820870X₇±557.93;④预测青春期女性肺活量正常参考值运用人工神经网络模型,构建5层神经网络,训练次数为3000次时,模型最优;⑤预测青年男性肺活量正常参考值运用非线性回归模型,其模型为:（?）=4890.41-0.03057X₁-0.094425X₂+27.2634X₃-1.3680X₃²-12.4397X₄-0.1151X₅-3381.2805/X₆+18.1272X₇±255.69;⑥预测青年女性肺活量正常参考值运用偏最小二乘回归模型,其模型为:（?）=2692.70-0.001604X₁+0.017849X₂-2.28365X₃-1.42546X₄-0.029721X₅+1.94182X₆+12.64521X₇±114.90;⑦预测中年男性肺活量正常参考值运用非线性回归模型,其模型为:（?）=-1702.56+0.2226X₁+0.1971X₂-15.6680X₃+48.4351X₄-0.4567X₅+153.8167X₆-3.1040X₆²+109.3521X₇±350.03;⑧预测中老年女性肺活量正常参考值运用多元线性回归模型,其模型为:（?）=3105.831-20.085X₆±230.427;⑨预测老年男性肺活量正常参考值运用非线性回归模型,其模型为:（?）=3184.086+171213.7967/X₅±300.33。应用以上的模型,代入了已选择4383个观测点的相应的地理因素指标,可计算出这4383个地方不同年龄不同性别的肺活量正常参考值,借助GIS空间分析中的地统计分析模块,通过克里格（Kriging）插值法精确的内插出肺活量正常参考值的空间趋势分布图。如果知道了中国某地的地理因素,就可用此模型估算该地区不同年龄不同性别的肺活量正常参考值,从地理分布图也可得到中国任何地方的正常参考值。更多还原

【Abstract】 Normal reference value of the vital capacity is a very important index on lung function testing.It can be used to diagnose the degree of restrictive impairment of respiratory function.The vital capacity increases or reduces is always explained by pathology or physiology which is much about human’s height,weight and age,ignoring the effect of environment factors such as topography,climate,and physiognomy on human’s lung function.After the full retrieval,it is discovered that although many papers have reported the relationship between normal value of vital capacity and geographical factors,only the qualitative description or the simple linear regression analysis is used.By far,no paper does such quantitative,linear and nonlinear research on the normal value of vital capacity and geographical factors based on the entire geographical environment and systemic study on China as a large region.The innovation of this paper is that normal value of vital capacity of the different age groups and the different gender groups of Chinese and geographical factors are quantitatively researched by linear and nonlinear methods,and deepening people’s cognition.To explore the complex relationship between normal reference value of vital capacity and seven geographical factors,through searching a lot of documents and buying data from some units,normal reference value of vital capacity of the different age and the different gender groups of people in all the Chinese regions can be collected.According to sharing materials from The State Bureau of Surveying and Mapping,The Chinese Meteorological Data Central,the related geographic works, dictionary and literature,altitude（X₁）,annual sunshine duration（X₂）,annual mean air temperature （X₃）;annual mean relative humidity（X₄）,annual precipitation amount（X₅）,annual range of air temperature（X₆）;annual mean wind speed（X₇） can be collected.This paper builds many models between vital capacity and geographical factors by multiply liner regression analysis,partial least squares regression analysis,ridge regression analysis,nonlinear regression analysis,BP artificial neural networks,compares the linear and the nonlinear models,then the best model is chosen to predict the vital capacity of the different age and the different gender groups of people in all the Chinese regions.The models are as follows:①Applying BP artificial neural networks,5 NN network is built to predict normal reference value of the vital capacity of boy,training time is 264,model is best;②Applying partial least squares regression analysis to predict normal reference value of the vital capacity of girl,the model is: （?）=1321.21-0.043514X₁+0.010103X₂-0.25246X₃-0.75417X₄-0.013527X₅ +3.83186X₆+38.61289X₇±133.50:③Applying partial least squares regression analysis to predict normal reference value of the vital capacity of adolescent boys,the model is:（?）=3024.77-0.131640X₁+0.022388X₂+2.381548X₃-1.344117X₄-0.030850X₅ +7.881706X₆+39.820870X₇±557.93:④Applying BP artificial neural networks,5 NN network is built to predict normal reference value of the vital capacity of adolescent girls,training time is 3000,model is best;⑤Applying nonlinear regression analysis to predict normal reference value of the vital capacity of young men,the model is:（?）=4890.41-0.03057X₁-0.094425X₂+27.2634X₃-1.3680X₃²-12.4397X₄ -0.1151X₅-3381.2805/X₆+18.1272X₇±255.69:⑥Applying partial least squares regression analysis to predict normal reference value of the vital capacity of young women,the model is:（?）=2692.70-0.001604X₁+0.017849X₂-2.28365X₃-1.42546X₄-0.029721X₅ +1.94182X₆+12.64521X₇±114.90:⑦Applying nonlinear regression analysis to predict normal reference value of the vital capacity of presenile men,the model is:（?）=-1702.56+0.2226X₁+0.1971X₂-15.6680X₃+48.4351X₄-0.4567X₅ +153.8167X₆-3.1040X₆²+109.3521X₇±350.03:⑧Applying multiply liner regression analysis to predict normal reference value of the vital capacity of middle-aged and old women,the model is:（?）=3105.831-20.085X₆±230.427:⑨Applying nonlinear regression analysis to predict normal reference value of the vital capacity of old men,the model is:（?）=3184.086+171213.7967/X₅±300.33Applying above models,geographical factors of 4383 observation points can be calculated to predict normal reference value of the vital capacity of the different age and the different gender groups of Chinese.By the method of geostatistical analysis module of GIS spatial analysis and Kriging interpolation,geographical distribution map of the normal reference value of the vital capacity is precisely interpolated.Therefore,if the value of geographical factors of some place in China is known,normal reference value of the vital capacity of the different age and the different gender groups of people can be estimated,and normal reference value oftbe vital capacity any place can be also directly obtained from geographical distribution map.更多还原