北京大学学报（医学版）
北京大學學報（醫學版）
북경대학학보（의학판）
JOURNAL OF BEIJING MEDICAL UNIVERSITY(HEALTH SCIENCES)
2010年
2期
197-201
,共5页
胃肠道%微球体%神经网络(计算机)%藻酸盐%阿司匹林
胃腸道%微毬體%神經網絡(計算機)%藻痠鹽%阿司匹林
위장도%미구체%신경망락(계산궤)%조산염%아사필림
Gastrointestinal tract%Microspheres%Neural networks(computer)%Alginates%Aspirin
目的:优化阿司匹林海藻酸钙胃漂浮缓释微球的制备方法.方法:以处方参数为输入、药物释放度和漂浮率为输出建立模型,同时运用人工神经网络和响应曲面法预测了药物的体外释放,优化了处方,并比较了两种方法的预测和优化性能.结果:人工神经网络预测的准确性更好,两种方法的优化结果相似.优化后,大约90%(体积分数)的微球可以在体外人工胃液中漂浮4 h以上,阿司匹林60 min内释放42.12%,120 min内释放60.97%,240 min内释放78.56%,药物释放符合Higuchi方程.结论:运用人工神经网络和响应曲面法成功制备了体外释放良好的阿司匹林胃漂浮微球.
目的:優化阿司匹林海藻痠鈣胃漂浮緩釋微毬的製備方法.方法:以處方參數為輸入、藥物釋放度和漂浮率為輸齣建立模型,同時運用人工神經網絡和響應麯麵法預測瞭藥物的體外釋放,優化瞭處方,併比較瞭兩種方法的預測和優化性能.結果:人工神經網絡預測的準確性更好,兩種方法的優化結果相似.優化後,大約90%(體積分數)的微毬可以在體外人工胃液中漂浮4 h以上,阿司匹林60 min內釋放42.12%,120 min內釋放60.97%,240 min內釋放78.56%,藥物釋放符閤Higuchi方程.結論:運用人工神經網絡和響應麯麵法成功製備瞭體外釋放良好的阿司匹林胃漂浮微毬.
목적:우화아사필림해조산개위표부완석미구적제비방법.방법:이처방삼수위수입、약물석방도화표부솔위수출건립모형,동시운용인공신경망락화향응곡면법예측료약물적체외석방,우화료처방,병비교료량충방법적예측화우화성능.결과:인공신경망락예측적준학성경호,량충방법적우화결과상사.우화후,대약90%(체적분수)적미구가이재체외인공위액중표부4 h이상,아사필림60 min내석방42.12%,120 min내석방60.97%,240 min내석방78.56%,약물석방부합Higuchi방정.결론:운용인공신경망락화향응곡면법성공제비료체외석방량호적아사필림위표부미구.
Objective:To investigate the preparation and optimization of calcium alginate floating micro-spheres loading aspirin.Methods:A model was used to predict the in vitro release of aspirin and optimize the formulation by artificial neural networks (ANNs) and response surface methodology(RSM).The amounts of the material in the formulation were used as inputs,while the release and floating rate of the mi-crospheres were used as outputs.The performances of ANNs and RSM were compared.Results:ANNs were more accurate in prediction.There Was no significant difierence between ANNs and RSM in optimiza-tion.Approximately 90% of the optimized microspheres could float on the artificial gastric juice over 4 hours.42.12% of aspirin was released in 60 min,60.97% in 120 min and 78.56% in 240 min.The re-lease of the drug from the microspheres complied with Higuchi equation.Conclusion:The aspirin floating microspheres with satisfying in vitro release were prepared successfully by the methods of ANNs and RSM.
