建立了一个设计二氢吡啶类钙拮抗剂骨架片的人工神经网络，该系统由输入层、隐含层、输出层组成，输入层有４个节点（药物的溶解度、骨架材料的水化速率和用量、表面活性剂的用量），隐含层有３个节点，输出层有３个节点（第２、４、６ｈ药物的释放量）。以卡波姆为骨架材料研究了该系统在设计硝苯地平与尼莫地平骨架片的可行性。结果表明：本网络能较好地模拟给定处方的药物释放情况 An artificial neural network for designing dihydropyridine calcium antagonist matrix was established. The system consisted of input layer, hidden layer and output layer. The input layer had 4 nodes (solubility of drug, hydration rate of matrix material and Dosage, amount of surfactant), there are 3 nodes in the hidden layer and 3 nodes in the output layer (the release amount of the 2nd, 4th, 6th drug). The feasibility of using this system to design nifedipine and nimodipine matrix tablets was investigated using carbomer as the framework material. The results show that this network can better simulate the drug release in a given prescription