【摘要】 目的观察地理环境与左心室短轴缩短率参考值的关系及其影响机制,为制定该参考值的统一标准提供更全面的科学依据。方法收集全国49个市县的2 252例健康中老年人左心室短轴缩短率的参考值,运用全局空间自相关的Moran’s I指数探索该医学指标与空间位置之间的关系,再利用相关分析方法,探寻该医学指标与9项地理因素的关系,进一步构建回归模型,分别构建基于遗传算法和网格搜索法的支持向量回归机。结果研究结果表明,中国中老年人左心室短轴缩短率参考值在空间上呈正的自相关性,并与其中5项地理指标具有显著的相关关系。经过比较分析,基于遗传算法的支持向量回归模型在预测精度上优于基于网格搜索法的模型。结论若已知某一地区有关地理要素的值,则可通过建立支持向量回归模型得出该区的中老年人左心室短轴缩短率的参考值。更多还原

【Abstract】 Objective To examine the relationship between geographical factors and left ventricular fractional shortening(LVFS)and the influence mechanism of LVFS in order to provide more complete scientific evidence for establishing the uniform standard for LVFS reference value.Methods A total of 2 252 LVFS values of healthy middle-aged and elderly people were collected from 49cities/counties of China.The relationship between the LVFS and space location was explored by using the method of spatial autocorrelation analysis based on the Moran statistics.The correlation analysis between the LVFS and 9geographical factors was performed.Furthermore,the support vector regression model was established based on genetic algorithm and grid search,respectively.Results The reference values of LVFS of middle-aged and elderly people had a positive spatial autocorrelation and they were significantly correlated with 5geographical factors.The support vector regression model based on genetic algorithm was superior to that on grid search method in terms of predictive accuracy.Conclusion The reference value of LVFS of middle-aged and elderly people can be determined based on the support vector regression model,if the geographical factors of a certain region are obtained.更多还原