Objective To study the correlation between geographical factors and cardiac index and to supply a basis for standardizing the normal reference value of Chinese elderly Women's cardiac index (CI) . Methods The correlation analysis was made between 2 135 healthy elderly Women'CI normal reference values of and geographical factors, which were determined by different hospitals, related units and institutions in 44 regions, and simulating the nonlinear relationship between them by BP neural network method. Results The correlation between the reference value of Chinese elderly healthy Women's CI and geographical factor was significant, the prediction value of CI with BP network had much higher accuracy, and the geographical distribution map of the normal reference value of Chinese elderly Women's CI was obtained by Kriging method. Conclusions BP neural networks can be used to well simulated the nonlinear relationship between CI and geographical factors, and the normal reference value of Chinese elderly healthy Women's CI in any place can be obtained from the geographical distribution map.%目的 探讨心脏指数和地理因素之间的相关性,为制定心脏指数正常参考值的统一标准提供科学依据.方法 收集了全国44个市、县、区级医院和有关单位及高等院校测定的2 135例正常健康老年女性心脏指数参考值,对其与地理因素之间的相关性进行显著性分析,并运用BP神经网络模拟其与地理因素之间的非线性关系.结果 发现中国健康正常老年女性的心脏指数与地理指标之间的相关性很显著,并且用BP网络得到的预测值有较高的精度,从而用克里格(Kriging)插值法精确的内插出中国老年女性心脏指数正常参考值的地理分布图.结论 BP神经网络可以很好地模拟心脏指数和地理因素之间的非线性关系,同时也可以从地理分布图得到中国任何地方老年女性心脏指数的正常参考值.
展开▼
