论文部分内容阅读
目的:将人工神经网络应用于环孢素A(CsA)剂量预测研究中,构建CsA剂量预测模型,优化个体化给药方案,提高临床药物治疗效果、减少用药不当引起的疾病。方法:收集了某三甲医院使用环孢素A的数据,利用MAT-LAB软件,使用遗传算法对BP网络权值、阈值进行优化,再用BP算法训练网络,进行拟合。结果:此网络的拟合度能达到99.428%,证明了可以利用此网络来进行预测,可以给临床医师的用药作为参考。
OBJECTIVE: To apply Artificial Neural Network (ANN) in CsA dose prediction study, construct CsA dose prediction model, optimize individualized drug delivery scheme, improve the effect of clinical drug treatment, and reduce the diseases caused by inappropriate medication. Methods: The data of cyclosporin A in a top three hospital were collected. The genetic algorithm was used to optimize the weights and thresholds of the BP network by MAT-LAB software. The BP network was trained and fitted to the network. Results: The fitting rate of this network can reach 99.428%, which proves that this network can be used for prediction. It can be used as a reference for clinicians.