以健康老年男性血沉参考值为例,收集中国各地用温氏法测定的健康老年男性血沉参考值,运用BP神经网络模拟分析其与地理环境的海拔高度、年日照时数、年平均相对湿度、年平均气温、年降水量等要素间的非线性、复杂性关系特征,通过5层神经网络,经过500次及1 300次自学习建立模拟规则后,用此规则很好地模拟健康老年男性血沉参考值与地理环境的关系。知道中国某地的地理要素, 就可用此模型估算该地区老年男性血沉参考值。人工神经网络预测血沉参考值的模型自动定义具体的结构和参数,减少血沉参考值预测中的人为性,计算方便;同时,通过对不同地理要素的组合,模型可以寻找预测误差最低的地理要素组合关系。
This paper discusses the nonlinear relationship between reference value (wintrobe) standard of Chinese old men's ESR and geography environments based on artificial neural network. Reference value (wintrobe) standard of Chinese old men's ESR is connected with geographical environment, diet structure and settlement environment, etc. which interact with blood indirectly. It is artificial neural network that have the ability to obtain the nonlinear relationship between variables. In order to supply a basis for uniting the reference value (wintrobe) standard of Chinese old men's ESR, a research is made about the nonlinear relationship between the reference value (wintrobe) of Chinese healthy old men's ESR and five geographical factors, which are determined by the way of wintrobe laws based on artificial neural network. After building 5-layer NN, the relationship was simulated by artifical neural network(NN). After training 500 times and 1300 times, the network works better. If the geographical factors known, the reference value (wintrobe) of Chinese old men's ESR can be simulated by the neural network more accurately.
[1] 裴澍萱,黄颖程,李成松,等.高原内科专题讲座[M].北京:人民卫生出版社,1993.
[2] 孔俭,冯影,高建立,等.长春地区血液流变学指标正常值的测定[J].中华医学检验杂志,1991,14(5):307～308.
[3] 郑志学,陈国华,梁子钧,等.健康老年人血液流变学[J].中华老年医学杂志,1982,1(2):101～104.
[4] 高佩琦,乔觉民,朱翔,等.黑龙江省入群的血液流变学指标正常值测定[J].哈尔滨医科大学学报,1993,27(1):7～8.
[5] 张必全,王敬华.118例健康成人血液流变学八项指标调查报告[J].临床检验杂志,1991,9(1):13.
[6] 孙志新,恽寿全,潘卫红,等.西宁、杭州两地老年人血液流变学的改变及其与中青年的比较[J].高原医学杂志,1994,4(4):7～10
[7] 何作云,尹翠娥,丁秋华.人体血液流变特性正常范围及其机理的初探[J].中国血液流变学杂志,1992,2(1):16～21.
[8] 杨春,许汉林,雷一凡,等.汉中地区1000名正常人血液流变学检测参考值调查报告[J].中国血液流变学杂志,1992,2(2):23～25.
[9] 许维勤,孟宪君.344例健康人血液流变学指标正常值测定[J].中国血液流变学杂志,1993,3(1):32～33.
[10] 葛淼,肖燕飞,刘昆,等.老年前期男性血红蛋白正常参考值与中国地理因素[J].地理科学,2004,24(6):767～770.
[11] Ge Miao,Yang Qingsheng,Ren Zhiyuan,et al.Reference value of Presenile human hematocrit and geographical factors[J].Journal of Clinical Laboratory Analysis,2002,16 (1):26-29.
[12] 葛淼,郭彩玲,南秀琴,等.血沉参考值(温氏法)与中国地理因素的关系[J].地理研究,1996,15(3):85～89.
[13] 马晓丽,宁丽娟,刘倩,等.青年男性血沉正常参考值与地理因素的关系[J].山西大学学报(自然科学版),2003,26(4):368～372.
[14] 翁维良,王怡,王仰宗,等.实用临床血液流变学[M].北京:学苑出版社,1994.
[15] 阎崇年,颜吉鹤,宋俊岭,等.中国市县大辞典[M].北京:中共中央党校出版社,1991.
[16] 赵济,陈永文,韩渊丰,等.中国自然地理(第三版)[M].北京:高等教育出版社,1995.1～110.
[17] 张超,杨秉赓.计量地理学基础(第二版)[M].北京:高等教育出版社,1991.
[18] 董长虹.Matlab神经网络及应用[M].北京:国防工业出版社,2005.
[19] Pattie D C,Snyder J.Using a neural network to forecast visitor behavior[J].Annuals of tourism research,1996,23 (1):151-164.
[20] 王士同.问题求解的人工智能:神经网络方法[M].北京:气象出版社,1995.
[21] Karayiannis N B,Venetsanopoulos A N.Artificial neural network:learning algorithems,Performance Evaluation and application[M].Boston:kluwer academic,1993.
[22] 张伟力,杜文经,何莹.温氏血沉在不同温度时的差异性分析[J].中国血液流变学杂志,2004,14(3):394～395.
