径向基神经网络预测利培酮的稳态血药浓度  

Prediction of Steady-state Blood Concentration of Risperidone Using RBF Neural Networks

作　　者：刘朝晖[1,2] 梅全喜[3] 黄榕波[4] 温预关[5] 李明亚[2] 

机构地区：[1]广东中山市中医院药剂科,广东中山528400 [2]广东药学院药科学院,广州510006 [3]广东中山市中医院科教科,广东中山528400 [4]广东药学院基础学院,广州510006 [5]广州市脑科医院/国家药品临床研究基地,广州510370

出　　处：《中国药房》2012年第26期2422-2424,共3页China Pharmacy

基　　金：国家自然科学基金资助项目(10926191);中山市科技计划资助项目(20102A024)

摘　　要：目的:评价用径向基(RBF)神经网络所建立的预测利培酮稳态血药浓度模型的预测性能。方法:将数据分为训练集、校验集和测试集来建立获取输出变量与输入变量两者间关系的RBF网络模型,其中以患者的性别、年龄、体重、剂量、血压、多项生理生化指标等37项参数为输入变量,利培酮稳态血药浓度为输出变量。用训练集和校验集的网络计算输出值与目标输出值之间的均方差(MSE)和相关系数(R)来综合评价网络模型的学习效果,用测试集的网络计算输出值与目标输出值之间的MSE和R来评价网络模型的预测性能。结果:当扩展系数值为1.5时,训练集的MSE为6.93×10-6,R值为0.99988;校验集的MSE为8.24×10-3,R值为0.86669;测试集的MSE为8.58×10-3,R值为0.80899;网络模型的预测效果和泛化能力较好。结论:RBF网络用于预测利培酮稳态血药浓度的研究是可行的。OBJECTIVE： To evaluate the performance of a model for predicting the steady-state blood concentration of risperidone established by using radial basis function （RBF） neural network. METHODS： The data was divided into training set, validation set and test set to establish the RBF neural network model which had captured the relationships between the input variables （37 parameters such as the patients＇ gender, age, weight, dosage, blood pressure and multiple physiological and biochemical indexes etc.） and the output variable （steady-state blood concentration of risperidone）. Learning effect of the model was comprehensively evaluated by error of mean square （MSE） and coefficient correlation （R） between the computed output value and objective output value of training set and validation set. And predictive performance of the model was evaluated by MSE and R between the computed output value and objective output value of test set. RESULTS： When the Spread value was 1.5, the MSE and R values of the training set, validation set and test set were 6.93×10%-6 and 0.999 88, 8.24×10^-3 and 0.866 69, 8.58×10^-3 and 0.808 99, respectively. The RBF neural network model had the better predictive effect and generalization. CONCLUSION： It is practical and valid for RBF neural network model to be applied to predict steady-state blood concentration of risperidone.

关 键 词：径向基神经网络 利培酮 稳态血药浓度 

分 类 号：R969.1[医药卫生—药理学] R971[医药卫生—药学]