Bagging偏最小二乘和Boosting偏最小二乘算法的金银花醇沉过程近红外光谱定量模型预测能力研究

建立金银花醇沉过程中稳健的近红外光谱( Near infrared spectroscopy,NIR)定量模型,为金银花醇沉过程的快速评价提供方法。研究基于金银花醇沉过程绿原酸的 NIR 数据,通过建立 Bagging 偏最小二乘(Bagging-PLS)模型、Boosting偏最小二乘(Boosting-PLS)模型与偏最小二乘(Partial Least Squares,PLS)模型,实现对模型性能比较;在此基础上,采用组合间隔偏最小二乘法( Synergy interval partial least squares,siPLS)和竞争自适应抽样( Competitive adaptive reweighted sampling,CARS )法分别对光谱进行变量筛选,建立模型,实现了对模型预测性能的考察。实验结果表明, Bagging-PLS和Boosting-PLS(潜变量因子数设为10)的预测性能均优于 PLS 模型。在此基础上,两批样品采用 siPLS 筛选变量,第一个批次金银花筛选波段820~1029.5 nm和1030~1239.5 nm,第二个批次金银花醇沉筛选波段为820~959.5 nm和960~1099.5 nm;采用CARS方法变量筛选,两批样品分别选择5折交叉验证和10折交叉验证,取交叉验证均方根误差( RMSECV)值最小的子集作为最终变量筛选的结果。经过变量筛选的两批金银花醇沉过程中的绿原酸含量Bagging-PLS和Boosting-PLS模型的预测均方根误差(RMSEP)值降低了0.02~0.04 g/L,预测相关系数提高了4%~5%。综上,Baggning-PLS和Boosting-PLS算法可作为金银花醇沉过程NIR定量模型的快速预测方法。     

过程分析技术(PAT)在原料药生产中的应用

该文通过采用近红外光谱分析技术对原料药(API)的浓度调节过程进行实时监控,介绍了在良好生产规范条件下过程分析技术(PAT)的实施过程。利用偏最小二乘算法开发出两个校正模型分别用以监控原料药和水分含量,并通过模型校正均方根误差(RMSEC)、交叉检验均方根误差(RMSECV)和预测均方根误差(RMSEP)以及对应的决定系数(R~2)来评估模型的性能。为保证模型性能,按照分析方法验证要求对模型的线性和范围、准确性、精密度(重复性)、专属性以及稳健性指标进行验证。最后通过系统性能测试确认检测系统满足商业化运行的要求。结果显示,采用过程分析技术控制浓度调节过程,可以大幅度缩短浓度调节时间,节约蒸汽能耗和检测费用,减少生产过程中的偏差,提升产品工艺水平和批次间一致性。     

刺五加注射液近红外含量预测模型谱段选择规律和消除溶剂干扰方法的探讨

目前中药制剂生产过程中缺乏全过程参数检测和质量控制技术手段,不同生产批次药品化学成分差异较大、质量不够稳定、临床使用疗效和安全性不理想,因此,建立其完善的质量评价体系及其准确快速的质量评价方法,成为中药质量控制的重中之重.通过对刺五加注射液近红外(near infrared,NIR)含量预测模型的谱段选择规律和消除溶剂干扰方法的探讨,发现采用表征混合物结构差异的结构相关谱段结合含量相关谱段作为NIR组分预测模型谱段,用基于水为参比光谱的样本光谱建立含量预测模型,并利用水作为参比光谱识别和提取待分析组分的光谱信息,可以得到较理想的NIR含量预测结果.通过对刺五加注射液中绿原酸、紫丁香苷和刺五加苷E组分的近红外光谱结构相关谱段和含量相关谱段的归属,分别建立了绿原酸、紫丁香苷和刺五加苷E组分的含量预测模型,可用于快速分析刺五加注射液中不同组分的含量.     

近红外光谱快速测定红花逆流提取过程中羟基红花黄色素A的含量

采用近红外光谱(NIRS)透射法对红花罐组式逆流提取过程中羟基红花黄色素A(Hydroxysafflor yellow A,HSYA)的含量进行快速无损的测定.在红花逆流提取过程中,以高效液相色谱法(HPLC)为对照分析方法,测定提取液中羟基红花黄色素A的含量,运用偏最小二乘(PLS)法建立NIR光谱与羟基红花黄色素A的HPLC分析值之间多元校正模型,并对逆流提取过程的未知样本进行含量预测.校正模型相关系数达到0.982,预测相关系数达到0.965,RMSEC和RMSEP分别为0.053和0.075,RSEC和RSEP分别为3.96%和5.25%.结果表明,NIRS可以作为一种准确、快速、无损的检测方法用于检测中药逆流提取过程有效成分含量变化规律.     

多批次肝衰竭患者呼出气体的电喷雾萃取电离质谱检测及代谢组学数据分析

采用高分辨电喷雾萃取电离质谱(EESI-MS)技术对肝衰竭患者和健康志愿者呼出气体样本进行快速检测, 结合多块偏最小二乘分析(MB-PLS)方法, 对多批次获取的呼出气体代谢数据进行统计建模分析, 并与传统的PLS方法进行比较. 结果表明, MB-PLS方法能有效消除批次差异对统计建模的影响. 此外, 利用MB-PLS模型变量VIP值对变量进行筛选, 可降低数据的冗余, 消除无关变量对模型的影响, 从而有效提高了模型的性能.     

中药琼玉膏中梓醇含量检测方法

摘要：发明属于中药技术领域。本发明提供了中药琼玉膏中梓醇含量的检测方法。本发明以君药地黄为基础，通过正交设计方法，     

人工神经网络-近红外光谱法用于甲氧苄胺嘧啶粉末药品的非破坏定量分析

近红外(NIR)光谱分析技术已应用于制药、化妆品、烟草、食品、化学药品、聚合物、纺织品、油漆涂料、 煤炭和石油工业等各个领域的质量监控. 近年来, NIR光谱分析技术也应用于药品分析中, 因该方法具有非破坏性, 样品不需要复杂的预处理和分离即可直接测定. 它可对药物进行定性和定量测定以及多晶、光学异构体和湿度的测定. 近红外光谱法用于无损非破坏测定胶囊类以及片剂的研究已有报道[1,2].     

太赫兹技术融合人工智能在中药质量控制中的应用和展望北大核心CSCD

中药是中华民族的瑰宝,中药的质量控制对于确保中药安全性和有效性至关重要。传统的中药质量控制方法费时费力,需要对样品进行复杂的前处理。太赫兹技术是一种新兴的无损检测技术,具有高分辨率、非接触、非破坏等优点,与人工智能算法结合,能够实现对中药生产全过程的质量控制。该文首先介绍了太赫兹技术的原理、太赫兹光谱预处理方法、特征数据提取方法、人工智能算法建模以及模型的评估与优化;并在此基础上,对近年来太赫兹技术在中药鉴别、含量测定和制药生产过程质量控制中的应用进行了综述。最后,对太赫兹技术的发展趋势及其在中药智能制造中的应用进行了展望。     

人工神经网络—近红外光谱法用于甲氧苄胺嘧啶粉末药品的非？…

近红外(NIR)光谱分析技术已应用于制药、化妆品、烟草、食品、化学药品、聚合物、纺织品、油漆涂料、煤炭和石油工业等各个领域的质量监控.近年来,NIR光谱分析技术也应用于药品分析中,因该方法具有非破坏性,样品不需要复杂的预处理和分离即可直接测定.它可对药物进行定性和定量测定以及多晶、光学异构体和湿度的测定.近红外光谱法用于无损非破坏测定胶囊类以及片剂的研究已有报道[1,2].NIR光谱在使用中也有一定的局限性,主要是结构复杂,谱图重叠多,在进行定性和定量分析中需采用一定的数据处理才能获得准确可靠的分析结果.在定量分析中,…     

一种基于最小二乘支持向量机算法的近红外光谱判别分析方法

将最小二乘支持向量机(LSSVM)用于近红外(NIR)光谱分析,建立一种新型的NIR光谱快速鉴别方法。以丹参药材道地性鉴别为例,对其NIR漫反射光谱进行主成分分析后,运用LSSVM法建立NIR光谱非线性分类模型,对丹参药材道地性进行快速鉴别。将本方法与经典SVM和BP神经网络法相比较,结果表明,本法判别准确率高,计算时间少,可推广应用于中药等天然产物质量快速鉴别。     

Self-modeling curve resolution (SMCR) analysis of near-infrared (NIR) imaging data of pharmaceutical tablets

The idea of quality by design (QbD) has been proposed in pharmaceutical field. QbD is a systematic approach to control the product performance based on the scientific understanding of the product quality and its manufacturing process. In the present study, near-infrared (NIR) imaging is utilized as a tool to achieve this concept. A practical use of a chemometrics technique called self-modeling curve resolution (SMCR) is demonstrated with NIR imaging analysis of pharmaceutical tablets containing two ingredients, a soluble active ingredient, pentoxifylline (PTX), and an insoluble excipient, palmitic acid. Concentration profiles obtained by SMCR reveal that the homogenous distribution of chemical ingredients strongly depends on the grinding time and that its process plays a central role in quantitative control, say sustained-release of PTX. In addition, pure component spectra by SMCR indicate a sequential change of specific NIR peak intensities following the increase of the grinding time. The spectra change shows a molecular structure change related to its crystallinity during grinding process. Accordingly, this study clearly demonstrates that NIR imaging combined with SMCR can be a powerful tool to reveal chemical or physical mechanism induced by the manufacturing process of pharmaceutical products and that it may be a solid solution for QbD of pharmaceutical products.     

Quality by Design applications in biosimilar pharmaceutical products

A process is well understood when all critical sources of variability are identified and explained, variability is managed by the process design and monitoring, and product quality attributes are accurately and reliably predicted over the design space. Quality by Design (QbD) is a systematic approach to product development and process control that begins with predefined objectives, emphasizes product and process understanding, and sets up process control based on sound science and quality risk management. The Food and Drug Administration (FDA) and the International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) have recently started promoting QbD in an attempt to curb rising development costs and regulatory barriers to innovation and creativity. QbD is partially based on the application of multivariate statistical methods and a statistical Design of Experiments strategy to the development of both analytical methods and pharmaceutical formulations. In this paper, we review the basics of QbD and their impact on the innovative, generic, and biosimilar pharmaceutical industry. In particular, we consider the challenge of mapping the control space in biotechnological processes and how advances in statistical methods can contribute to QbD.     

In-line monitoring of alcohol precipitation by near-infrared spectroscopy in conjunction with multivariate batch modeling

Alcohol precipitation is a critical unit operation during the manufacture of Chinese herbal injections. To facilitate enhanced process understanding and develop control strategy, the use of near-infrared spectroscopy (NIRS) combined with multivariate statistical process control (MSPC) methodology was investigated for in-line monitoring of alcohol precipitation. The effectiveness of the proposed approach was evaluated through an experimental campaign. Six batches were run under normal operating conditions to study batch-to-batch variation or batch reproducibility and establish MSPC control limits, while artificial process variations were purposefully introduced into the four test batches to assess the capability of the model for real-time fault detection. Several MSPC tools were compared and assessed. NIRS, in conjunction with MSPC, has proven to be a feasible process analytical technology (PAT) tool for monitoring batch evolution and potentially facilitating model-based advanced process control of the alcohol precipitation during the manufacture of Chinese herbal injections.     

Quality by Design: Design of Experiments Approach Prior to the Validation of a Stability-Indicating HPLC Method for Montelukast

This paper describes a systematic design of experiments (DoE) approach by applying the principle of quality by design (QbD) to determine the design space for a stability-indicating HPLC method prior to validation. By employing DoE, a simultaneous multivariate approach was carried out for mobile phase pH, flow rate, percentage of organic content and column temperature. A two-level fractional factorial design (2^4?1 + 2 center points = 10 experiments) was employed and statistical analysis of the experimental data uncovered the significant influential chromatographic factors. The experimental data for USP tailing and resolution were analyzed statistically to screen the chromatographic factors. This approach determined the most influential chromatographic factors. During this process, inferences were evaluated from various data tables, for example, analysis of variance, summary of fit, lack of fit, and parameter estimates. The study also explained various plots such as actual vs. predicted plot, Pareto plot, and prediction profiler. The acceptable range of the chromatographic factors was displayed as a Contour plot defining the ‘design space’ of the method. The range of operating conditions that guarantee a satisfactory QbD was deduced to finalize the method prior to validation. The method is simple, rapid, and robust for the determination of montelukast in montelukast sodium oral granules dosage form. The method was validated according to ICH guidelines for accuracy, precision, linearity, range, specificity, ruggedness and robustness (one factor varied at a time). The method has been successfully transferred to the quality control department for quality analysis of manufactured batches and stability samples.     

Quality by Design: Design of Experiments Approach Prior to the Validation of a Stability-Indicating HPLC Method for Montelukast

This paper describes a systematic design of experiments (DoE) approach by applying the principle of quality by design (QbD) to determine the design space for a stability-indicating HPLC method prior to validation. By employing DoE, a simultaneous multivariate approach was carried out for mobile phase pH, flow rate, percentage of organic content and column temperature. A two-level fractional factorial design (2⁴⁻¹ + 2 center points = 10 experiments) was employed and statistical analysis of the experimental data uncovered the significant influential chromatographic factors. The experimental data for USP tailing and resolution were analyzed statistically to screen the chromatographic factors. This approach determined the most influential chromatographic factors. During this process, inferences were evaluated from various data tables, for example, analysis of variance, summary of fit, lack of fit, and parameter estimates. The study also explained various plots such as actual vs. predicted plot, Pareto plot, and prediction profiler. The acceptable range of the chromatographic factors was displayed as a Contour plot defining the ‘design space’ of the method. The range of operating conditions that guarantee a satisfactory QbD was deduced to finalize the method prior to validation. The method is simple, rapid, and robust for the determination of montelukast in montelukast sodium oral granules dosage form. The method was validated according to ICH guidelines for accuracy, precision, linearity, range, specificity, ruggedness and robustness (one factor varied at a time). The method has been successfully transferred to the quality control department for quality analysis of manufactured batches and stability samples.

     

Analytical quality by design in the development of a cyclodextrin‐modified capillary electrophoresis method for the assay of metformin and its related substances

Quality by Design (QbD) is a new paradigm of quality to be applied to pharmaceutical products and processes, recently encouraged by International Conference on Harmonisation guidelines. In this paper QbD approach was applied to the development of a CE method for the simultaneous assay of metformin hydrochloride (MET) and its main impurities. QbD strategy was focused on electrophoretic process understanding, and the analytical method was thoroughly evaluated by applying risk assessment and chemometric tools. Method scouting allowed CD‐CZE based on the addition of carboxymethyl‐β‐CD to Britton‐Robinson acidic buffer to be chosen as operative mode. Seven critical process parameters (CPPs) were selected, related to capillary, injection, BGE and instrumental settings. The effect of the different levels of the CPPs on critical quality attributes (CQAs), e.g. critical resolution values and analysis time, was evaluated in a screening study. Response surface methodology led to draw contour plots and sweet spot plots. The definition of design space was accomplished by applying Monte‐Carlo simulations, thus identifying by risk of failure maps a multivariate zone where the CQAs fulfilled the requirements with a selected probability. Finally, a control strategy was designed and the method was applied to a real sample of MET tablets.     

Quality by Design approach in the development of a solvent-modified micellar electrokinetic chromatography method: Finding the design space for the determination of amitriptyline and its impurities

A solvent-modified micellar electrokinetic chromatography method was set up for the simultaneous determination of the tricyclic antidepressant amitriptyline (AMI) and its main impurities. The method was developed following Quality by Design (QbD) principles according to ICH Guideline Q8(R2). QbD approach made it possible to find the design space (DS), where quality was assured. After a scouting phase, aimed at selecting a suitable capillary electrophoresis pseudostationary phase, risk assessment tools were employed to define the critical process parameters (CPPs) to be considered in a screening phase (applied voltage, concentration and pH of the background electrolyte, concentration of the surfactant sodium dodecyl sulphate, of the cosurfactant n-butanol and of the organic modifiers acetonitrile and urea). The effects of the seven selected CPPs on critical quality attributes (CQAs), namely resolution values between critical peak pairs and analysis time, were investigated throughout the knowledge space by means of a symmetric screening matrix. Response surface study was then carried out on four selected CPPs by applying a Doehlert Design. Monte-Carlo simulations were performed in order to estimate the probability of meeting the desired specifications on CQAs, and thus to define the DS by means of a risk of failure map. Additional points at the edges of the DS were tested in order to verify the requirements for CQAs to be fulfilled. A control strategy was implemented by defining system suitability tests. The developed method was validated following ICH Guideline Q2(R1), including robustness assessment by Plackett–Burman design, and was applied to the analysis of real samples of amitriptyline coated tablets.     

Quality by design (QbD) based development of a stability indicating HPLC method for drug and impurities

In this paper, an application of Quality by Design (QbD) concepts to the development of a stability indicating HPLC method for a complex pain management drug product containing drug substance, two preservatives, and their degradants is described. The QbD approach consisted of (i) developing a full understanding of the intended purpose, (ii) developing predictive solutions, (iii) designing a meaningful system suitability solution that helps to identify failure modes, and (iv) following design of experiments (DOE) approach. The starting method lacked any resolution among drug degradant and preservative oxidative degradant peaks, and peaks for preservative and another drug degradant. The method optimization was accomplished using Fusion AE? software (S-Matrix Corporation, Eureka, CA) that follows a DOE approach. Column temperature (50 ± 5°C), mobile phase buffer pH (2.9 ± 0.2), initial % acetonitrile (ACN, 2 ± 1%), and initial hold time (2.5, 5, or 10 min) of the HPLC method were simultaneously studied to optimize separation of the unresolved peaks. The optimized HPLC conditions (column temperature of 50°C, buffer pH of 3.1, 3% initial ACN with 2.5 min initial hold) resulted in fully resolved peaks in the two critical pairs. The QbD based method development helped in generating a design space and operating space with knowledge of all method performance characteristics and limitations and successful method robustness within the operating space.     

Application of quality by design to the development of analytical separation methods

Recent pharmaceutical regulatory documents have stressed the critical importance of applying quality by design (QbD) principles for in-depth process understanding to ensure that product quality is built in by design. This article outlines the application of QbD concepts to the development of analytical separation methods, for example chromatography and capillary electrophoresis. QbD tools, for example risk assessment and design of experiments, enable enhanced quality to be integrated into the analytical method, enabling earlier understanding and identification of variables affecting method performance. A QbD guide is described, from identification of quality target product profile to definition of control strategy, emphasizing the main differences from the traditional quality by testing (QbT) approach. The different ways several authors have treated single QbD steps of method development are reviewed and compared. In a final section on outlook, attention is focused on general issues which have arisen from the surveyed literature, and on the need to change the researcher’s mindset from the QbT to QbD approach as an important analytical trend for the near future.