应用质量源于设计(Quality by Design，QbD)理念进行药品研发和生产，核心是确立关键工艺参数、关键物料属性等输入变量与关键质量属性等输出变量之间的多维组合与相互关系，并建立设计空间和控制策略。这一过程需要数学模型的辅助，以统计模型的应用最广。文章综述了实验设计、潜变量模型、人工神经网络和支持向量机4种常用统计模型的基本原理及其在基于QbD理念的药品研发与生产中的应用，以期为后续研究者选择合适的模型来确定输入-输出变量间的关系提供参考。
Application of Statistical Model in Drug Development and Production Based on QbD Concept
Applying the concept of Quality by Design(QbD)to drug development and production,the core is establishing the multidimensional combination and interrelationship between the input variables such as key process parameters and material attributes and the output variables such as key quality attributes,and establish the design space and control strategy.This process requires the aid of mathematical models,with statistical models being the most widely used.This paper reviews the basic principles of four commonly used statistical models,namely experimental design,latent variable model,artificial neural network and support vector machine,and their applications in drug development and production based on QbD concept,in order to provide reference for future researchers to select appropriate models to determine the relationship between input and output variables.
Quality by Designexperimental designlatent variable modelartificial neural networksupport vector machine
万方数据
维普
