For a long time, very few studies has referred to the evaluation of model applicability in near-infrared (NIR) quantitative analysis, which restricts the development and application of NIR technology.Our previous research demonstrated that common chemometric indicators, as applicability evaluation statistics of the NIR model, were only suitable for macroanalysis but not microanalysis. Due to these limitations, this project proposes a novel assessment method based on the fusion of multi-source information. Taking pharmaceutical raw materials, intermediates and final products of solid preparation (Angong Niuhuang Pill) and liquid preparation (Qingkailing Injection) as the research carriers, the project integrates three approaches: (1) evaluation indicators based on the accuracy profile method; (2) multivariate detection limits (MDL) and multivariate quantitation limits (MQL) baased on the Neyman-Pearson detection theory; (3) multivariable sensitivity analysis based on the net analysis signal method. The goal is to realize the multi-source information fusion of the NIR model evaluation indicators, including accuracy, trueness, precision, risk, linearity, MDL, MQL, uncertainty and sensitivity. It is believed that the implementation of the project will enhance the scientificity of the NIR model evaluation indicators for the quality control of Chinese material medica (CMM), as well as provide the innovative theory and methodology for the reasonable application of NIR in the process analysis of CMM.
近红外定量分析模型适用性评价研究的滞后,长期以来制约着近红外分析技术的应用和发展。课题前期基础研究工作显示,常规化学计量学指示参数仅适用于常量组分的近红外分析而非微量组分。针对此局限性,本项目提出多源信息融合的中药近红外模型适用性评价方法,以固体制剂(安宫牛黄丸)和液体制剂(清开灵注射液)原料药、生产过程中间体和成品为载体,采用基于准确性轮廓方法的模型适用性评价参数、基于"奈曼-皮尔逊"检测理论的多变量检测限和多变量定量限参数、以及基于净分析信号理论的多变量灵敏度参数三方面的技术途径,实现近红外模型适用性评价的准确性、真实性、精密度、风险性、线性、定量限、检测限、不确定性和灵敏性多源信息融合的目标,提高中药近红外模型适用性评价方法的科学性,为近红外分析技术在中药生产过程质量控制领域的合理运用提供创新方法学和创新理论。
近红外定量分析模型适用性评价研究的滞后,长期以来制约着近红外分析技术的应用和发展。课题前期基础研究工作显示,常规化学计量学指示参数仅适用于常量组分的近红外分析而非微量组分。针对此局限性,本项目提出多源信息融合的中药近红外模型适用性评价方法,以固体制剂(安宫牛黄丸)和液体制剂(清开灵注射液)原料药、生产过程中间体和成品为载体,采用基于准确性轮廓方法的模型适用性评价参数、基于"奈曼-皮尔逊"检测理论的多变量检测限和多变量定量限参数、以及基于净分析信号理论的多变量灵敏度参数三方面的技术途径,实现了近红外模型适用性评价的准确性、真实性、精密度、风险性、线性、定量限、检测限、不确定性和灵敏性多源信息融合的目标,提高中药近红外模型适用性评价方法的科学性,为近红外分析技术在中药生产过程质量控制领域的合理运用提供创新方法学。
{{i.achievement_title}}
数据更新时间:2023-05-31
玉米叶向值的全基因组关联分析
正交异性钢桥面板纵肋-面板疲劳开裂的CFRP加固研究
硬件木马:关键问题研究进展及新动向
基于SSVEP 直接脑控机器人方向和速度研究
小跨高比钢板- 混凝土组合连梁抗剪承载力计算方法研究
基于多源信息融合方法的中医诊疗指南临床应用多维评价模型研究
中药近红外定量模型跨尺度传递方法研究
多源信息融合的次任务驾驶安全性评价方法研究
中药过程质量控制的近红外光谱高性能模型融合