神经网络在预测难溶性药物从HPMC骨架片中释放的应用  

Application of neural network in predicting the release of insoluble drugs from HPMC matrix tablets

作　　者：叶婷 魏晓红[2] 梁文权[2] 吴建军[3] 

机构地区：[1]浙江省医药保健品进出口公司,浙江杭州310031 [2]浙江大学湖滨校区药学院,浙江杭州310031 [3]浙江大学理学院数学系,浙江杭州310031

出　　处：《中国药学杂志》2002年第10期758-760,共3页Chinese Pharmaceutical Journal

摘　　要：目的　运用神经网络BP模型对难溶性药物在HPMC亲水凝胶骨架片中的释放进行预测。方法　选取甲氧苄啶、卡马西平、磺胺甲唑、茶碱作为模型药物 ,考察不同pH值、不同离子强度的溶出介质对其释药的影响 ,并根据神经网络BP模型 ,借助计算机运算 ,预测不同药物在不同溶出介质条件下的释放。结果　用BP神经网络所建释药模型的拟合值、预测值均与实测值符合度较好。OBJECTIVE: The BP model of neural network was explored to predict the release of insoluble drugs from HPMC matrix tablets in different mediums. METHODS: Trimethoprim, carbamazepine, sulfamethoxazole and theophylline were selected as the poor soluble model drugs. The BP model of neural network with a mathematical formula was used to predict the drug release in the medium with different pH and ionic strength. RESULTS: The study showed that the mathematical model predicted the release effectively. CONCLUSION: The neural network can be used to precisely predict the release of insoluble drugs from HPMC matrix tablets in different medium.

关 键 词：神经网络 预测 难溶性药物 HPMC骨架片 羟丙基甲基纤维素 释放度 

分 类 号：R944[医药卫生—药剂学]