节点文献

基于人工神经网络的青年女性红细胞计数正常参考值的地理分布

Geographic distribution of normal reference value of young female’s red blood cell count based on artificial neural networks

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 崔致远葛淼曹利霞王欣何进伟梁伟

【Author】 CUI Zhi-Yuan,GE Miao,CAO Li-Xia,WANG Xin,HE Jin-Wei,LIANG WeiCollege of Tourism and Environment,Shaanxi Normal University,Xi’an 710062,Shaanxi,China

【机构】 陕西师范大学旅游与环境学院

【摘要】 目的:揭示中国青年女性红细胞计数正常参考值的地理分布,为制定其统一标准提供科学依据.方法:收集中国377个单位用显微镜技术法测定的38 263例青年女性红细胞计数正常参考值,运用BP神经网络和GIS空间分析,研究其与海拔高度、年日照时数、年平均相对湿度、年平均气温、年降水量等地理因素间的非线性、复杂性关系特征及红细胞计数正常参考值的地理分布.结果:通过5层神经网络,经过1800次自学习建立模拟规则后,用此规则很好地模拟健康青年女性红细胞计数正常参考值与地理因素的关系,并通过克里格插值(Kriging)法精确的内插出青年女性红细胞计数正常参考值的地理分布图.结论: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青年女性红细胞计数正常参考值,从地理分布图也可得到中国任何地方的正常参考值.

【Abstract】 AIM: To provide basis for unifying normal red blood cell(RBC) reference values of young women in various region of China.METHODS: The RBC reference value of 38263 young women determined with microscope technology from 377 medical units in various regions of China were collected.The nonlinear,complexity relationship characteristic between the RBC reference values and geographical factors including latitude,yearly mean sunshine hours,yearly mean relative humidity,yearly mean temperature and yearly precipitation,were analyzed by methods of BP Neural Network and Spatial analysis of GIS.RESULTS: The relationship between RBC reference values and geographical factors were well-imitated by the rules which was established after 1800 self-study by 5 layers’ Neural Networks,and exactly inserted the geography distribution figure of RBC reference values with methods of Kriging.CONCLUSION: If altitude of a particular area is known,the RBC reference value of Chinese young women in this area can be calculated using this model,and we can also get the RBC reference values of Chinese young women of anywhere in China from the geography distribution figure.

【基金】 国家自然科学基金(40671005);陕西师范大学研究生培养创新基金(2007CXS012)
  • 【文献出处】 第四军医大学学报 ,Journal of the Fourth Military Medical University , 编辑部邮箱 ,2008年14期
  • 【分类号】R188
  • 【被引频次】2
  • 【下载频次】120
节点文献中: 

本文链接的文献网络图示:

本文的引文网络