[Application scenario design and prospect of generative artificial intelligence (AI) in intelligent manufacturing and supply chain of traditional Chinese medicine].

Intelligent manufacturing technologies, including databases, mathematical modeling, and information systems have played a significant role in process control, production management, and supply chain management in traditional Chinese medicine(TCM) industry. However, their ability to process and utilize unstructured data, such as research and development reports, batch production records, quality inspection records, and supplier documents, is relatively weak. For text, images, language, and other unstructured data, generative artificial intelligence(AI) technology has shown strong potential for development in extracting information, extracting knowledge, semantic retrieval, and content generation. Generative AI is expected to provide a feasible set of tools for the utilization of unstructured data resources in the TCM industry. Based on years of research and industrial application experience in TCM intelligent manufacturing technology, this study reviewed the current situation of intelligent manufacturing in TCM and the utilization of unstructured data, analyzed the application value of generative AI in the TCM manufacturing process and supply chain, summarized four typical application scenarios, including intelligent pharmaceutical knowledge base/knowledge graph, intelligent on-the-job trai-ning, intelligent production quality control, and intelligent supply chain. Furthermore, this study also explained the data collection and processing, business process design, application potential, and value of each scenario based on industry demands. Finally, based on the integration of generative AI and TCM industrial models, the study proposed a preliminary concept of a smart industrial brain for TCM, aiming to provide a reference for the application of AI technology in the field of TCM manufacturing.

[Methodology and application of process analytical technology (PAT) for traditional Chinese medicine manufacturing:a review].

Owing to the advancement in pharmaceutical technology, traditional Chinese medicine industry has seen rapid development. Preferring conventional manufacturing mode, pharmaceutical enterprises of traditional Chinese medicine have no effective process detection tools and process control methods. As a result, the quality of the final products mainly depends on testing and the quality is inconsistent in the same batch. Process analytical technology(PAT) for traditional Chinese medicine manufacturing, as one of the key advanced manufacturing techniques, can break through the bottleneck in quality control of medicine manufacturing, thus improving the production efficiency and product quality and reducing the material and energy consumption. It is applicable to the process control and real-time release of advanced manufacturing modes such as intelligent manufacturing and continuous manufacturing. This paper summarized the general idea of PAT for traditional Chinese medicine manufacturing. Through the analysis of the characteristics and status quo of the technology, we summed up the methodology for the continuous application and improvement of PAT during the whole life-cycle of traditional Chinese medicine. The five key procedures(process understanding, process detection, process modeling, process control, and continuous improvement) were summarized, and the application was reviewed. Finally, we proposed suggestions for the technical and regulatory challenges in implementing PAT in traditional Chinese medicine industry. This paper aims to provide a reference for development and application of PAT in advanced manufacturing, intelligent manufacturing, and continuous manufacturing of traditional Chinese medicine industry.

[Application progress on data-driven technologies in intelligent manufacturing of traditional Chinese medicine extraction].

The application of new-generation information technologies such as big data, the internet of things(IoT), and cloud computing in the traditional Chinese medicine(TCM)manufacturing industry is gradually deepening, driving the intelligent transformation and upgrading of the TCM industry. At the current stage, there are challenges in understanding the extraction process and its mechanisms in TCM. Online detection technology faces difficulties in making breakthroughs, and data throughout the entire production process is scattered, lacking valuable mining and utilization, which significantly hinders the intelligent upgrading of the TCM industry. Applying data-driven technologies in the process of TCM extraction can enhance the understanding of the extraction process, achieve precise control, and effectively improve the quality of TCM products. This article analyzed the technological bottlenecks in the production process of TCM extraction, summarized commonly used data-driven algorithms in the research and production control of extraction processes, and reviewed the progress in the application of data-driven technologies in the following five aspects: mechanism analysis of the extraction process, process development and optimization, online detection, process control, and production management. This article is expected to provide references for optimizing the extraction process and intelligent production of TCM.

[Research on batch-to-batch quality control of compound Danshen extract based on physical characterization and multivariate statistical analysis].

Physical attributes of Chinese herbal extracts are determined by their chemical components, and the physical and chemical attributes jointly affect the preparation process performance and the final product quality. Therefore, in order to improve the quality control of Chinese herbal extracts, we should comprehensively study the batch-to-batch consistency of physical and chemical attributes as well as the correlations between them. This paper first explored the physical attributes affecting the preparation process performance of the compound Danshen extract and developed a method for characterizing the texture attributes. With such main chemical components as water, phenolic acids, saponins, and saccharides and texture, rheology, and other physical attributes taken into consideration, the batch-to-batch quality fluctuation of products from different production lines and time was analyzed by principal components analysis(PCA). Finally, the correlation and partial least squares(PLS) analysis was conducted, and the regression equation was established. The fitting result of the PLS model for dynamic viscosity was satisfying(R~2Y=0.857, Q~2=0.793), suggesting that the chemical components could be adjusted by the component transfer rate in the extraction process, the impurity removal rate in the alcohol precipitation process, and the water retention rate of the concentration process to meet the control of the extract dynamic viscosity. This study clarified the correlations between physical and chemical attributes of the compound Danshen extract and established a method for controlling its physical attributes based on process regulation, which would provide reference for improving the quality control of Chinese herbal extracts.

[Granulation of Yangxue Qingnao Granules in fluidized bed based on near-infrared spectroscopy].

To realize the real-time monitoring of the production process of Yangxue Qingnao Granules and improve the inter-batch consistency of granule quality in the granulation process, this study established a near-infrared quantitative prediction model of moisture, particle size, bulk density, and angle of repose in the fluidized bed granulation process of Yangxue Qingnao Granules based on near-infrared spectroscopy(NIRS). The near-infrared spectra were collected from 355 samples in 12 batches in the granulation process by integrating the sphere detection module of the near-infrared spectrometer. In combination with the pretreatment methods such as the first derivative, multiplicative scatter correction(MSC), and standard normal variate(SNV), the model was established by partial least squares(PLS) regression. The root mean square error of prediction(RMSEP) of moisture was 0.347 and R_P~2 was 0.935. The RMSEP of the D_(50) particle size model was 38.4 and R_P~2 was 0.980. The RMSEPs of bulk density and angle of repose were 0.018 8 and 0.879, with R_P~2 of 0.085 9 and 0.958. The results showed that the prediction of the PLS quantitative model combined with NIRS was accurate, and this model can be applied to the monitoring of key quality attributes in the fluidized bed granulation of Chinese medicinal granules in the production scale.

Establishment and validation of the quantitative analysis of multi-components by single marker for the quality control of Qishen Yiqi dripping pills by high-performance liquid chromatography with charged aerosol detection.

INTRODUCTION: Charged aerosol detection (CAD) has the merits of high sensitivity, high universality and response uniformity. The strategy that combines the quantitative analysis of multi-components by single marker (QAMS) with CAD has certain advantages for the quantification of multi-components. However, relevant research was limited. OBJECTIVES: To comprehensively investigate the crucial factors that affect the performance of the HPLC-CAD-QAMS approach and validate the credibility and feasibility of the method. METHODOLOGY: Multiple components of Qishen Yiqi dripping pills (QSYQ) were assayed using the high-performance liquid chromatography (HPLC)-CAD-QAMS approach. Some factors that affect the sensitivity and accuracy of the approach were sufficiently studied. After the method verification, principal component analysis (PCA) was applied to evaluate the quality consistency of three types of samples: normal samples, expired samples and negative samples. RESULTS: A HPLC-CAD-QAMS method was successfully developed for the multi-component determination of QSYQ. First, chromatographic conditions were optimised by a definitive screening design, and the optimised ranges of operating parameters were obtained with a Monte Carlo simulation method. Next, a new method to select the internal reference standards was successfully introduced based on the heatmap of Pearson correlation coefficients of the response factors. Then, the multi-point method was selected to calculate the relative correction factors, and a robustness test was conducted with Plackett-Burman design. Finally, the PCA was proved to be effective for the quality consistency evaluation of different samples. CONCLUSION: The developed HPLC-CAD-QAMS method can be a reliable and superior means for the multi-component quantitative analysis of QSYQ.

[Research on feedforward control method of mixing process for compound Danshen Dripping Pills based on quality by design(QbD)].

The mixing process is one of the key operation units for solid preparation of traditional Chinese medicine. The physical properties such as particle size, density and viscosity of the mixture are key factors that need to be controlled, which will directly affect the performance of the preparation molding process and product quality. Subsequent dripping process performance and appearance qua-lity of dripping pills will be affected by dynamic viscosity of materials in the mixing process. Based on this, with mixing process of compound Danshen dripping pills as the object, a feedforward control method for the dripping pill mixing process was established based on the concept of quality by design(QbD). Firstly, critical quality attribute(CQA)-dynamic viscosity, critical material attributes(CMAs)-the moisture content of compound Danshen extract, average molecular weight of polyethylene glycol 6000 and critical process parameter(CPP)-mixing temperature were identified through the analysis of properties for multiple batches of the raw materials and excipients as well as technological mechanism. Then the Box-Behnken experimental design was used to establish the regression model among CMA, CPP and CMA(R■=0.972 0, RMSE =16.24) to obtain the design space. Finally, through the verification of three batches within the design space, the mixing process temperature was adjusted according to the properties of the raw materials and exci-pients to achieve accurate control of the dynamic viscosity attribute. The relative deviation between the actual dynamic viscosity value and the target value was less than 3.0 %. The feedforward control of the mixing process of compound Danshen dripping pills was rea-lized in this study, which can contribute to improving quality consistency of the mixing process intermediates, simultaneously provide a reference for the research on the process quality control of other Chinese medicine dripping pills.

[Study of intelligent manufacturing quality digitalization of traditional Chinese medicine and practice on Compound Danshen Dripping Pills].

China healthcare industry has gradually developed the consumer-centric integrated service model. To satisfy consumers' increasing demands on pluralistic, personalized and transparent healthcare services, pharmaceutical manufacturing enterprises must provide high-quality, precise and flexible medicines. This can be achieved by accelerating implementation of intelligent manufacturing, which is the core competitiveness of pharmaceutical manufacturing enterprises. According to the authors' intelligent manufacturing projects in a traditional Chinese medicine(TCM) factory, study and industrial practice on intelligent manufacturing were presented in this paper. First, the quality digitalization-based intelligent manufacturing methodology of TCM was proposed in this paper. The methodology mainly included three digitalized technologies in process and quality design, manufacturing process control and product batch evaluation. Next, the architectural design of intelligent manufacturing systems in one TCM factory was introduced, and the functional modules and data transmission relationships covering seedling, cultivation, herbal slices, preparation, storage and quality management systems were described. Finally, these technologies were fully used, and an integrated quality digitalization system was successfully established in the production workshop of a TCM product Compound Danshen Dripping Pills. The actual operation and application of process analyzers, supervisory control and data acquisition(SCADA), manufacturing execution system(MES), data analysis system, and enterprise resource planning system(ERP) were introduced. This paper provides reference for technical path planning and systematic architecture of TCM intelligent manufacturing.

Batch-to-batch quality consistency evaluation of botanical drug products using multivariate statistical analysis of the chromatographic fingerprint.

Botanical drug products have batch-to-batch quality variability due to botanical raw materials and the current manufacturing process. The rational evaluation and control of product quality consistency are essential to ensure the efficacy and safety. Chromatographic fingerprinting is an important and widely used tool to characterize the chemical composition of botanical drug products. Multivariate statistical analysis has showed its efficacy and applicability in the quality evaluation of many kinds of industrial products. In this paper, the combined use of multivariate statistical analysis and chromatographic fingerprinting is presented here to evaluate batch-to-batch quality consistency of botanical drug products. A typical botanical drug product in China, Shenmai injection, was selected as the example to demonstrate the feasibility of this approach. The high-performance liquid chromatographic fingerprint data of historical batches were collected from a traditional Chinese medicine manufacturing factory. Characteristic peaks were weighted by their variability among production batches. A principal component analysis model was established after outliers were modified or removed. Multivariate (Hotelling T(2) and DModX) control charts were finally successfully applied to evaluate the quality consistency. The results suggest useful applications for a combination of multivariate statistical analysis with chromatographic fingerprinting in batch-to-batch quality consistency evaluation for the manufacture of botanical drug products.

[Multistage multivariate statistic quality control (MMSQC) for monitoring production process of traditional Chinese medicines].

OBJECTIVE: To establish a method for monitoring the quality of intermediates generated in each working procedure during the production process of traditional Chinese medicine (TCM) , in order to ensure the batch-to-batch quality consistency of TCM products. METHOD: The multistage multivariate statistic quality control (MMSQC) was proposed to monitor production quality of TCMs based on multivariate data analysis technique. Hotelling T2 and SPE were adopted for monitoring the quality of intermediates generated in each working procedure. Danshen injection was taken as the example to introduce the application method of MMSQC. RESULT: MMSQC can monitor the quality of intermediates generated in multiple working procedures, which is simpler and more accurate compared with single-indicator monitoring method. CONCLUSION: MMSQC can be popularized to monitor quality of multistage production of TCMs.

Development and validation of in-line near-infrared spectroscopy based analytical method for commercial production of a botanical drug product.

Near-infrared (NIR) spectroscopy is one of the most successful pharmaceutical process analytical tools. For botanical drug products (BDPs), many studies have applied NIR spectroscopy for rapid analysis of botanical raw materials, extracts and formulations. However, the real-time process analysis reported for BDPs was still mainly conducted on lab- or pilot- scale equipment, where some essential conditions of the materials and process parameters can be easily controlled for NIR spectral measurement. Due to the chemical and physical characteristics of the commercial production of BDPs, it is challenging to develop in-line NIR methods with enough robustness for industrial-scale application. This is the first reported case study of the development and validation of the in-line NIR method for the commercial production of BDPs, taking Compound Danshen Dripping Pill (CDDP) as an example. An in-line NIR method was developed for simultaneous measurement of the three critical quality attributes, i.e. the relative density, the moisture content and the content of 3,4-dihydroxyphenyl lactic acid (danshensu, a key active compound), during the extract concentration process. The NIR spectra and sample collection lasted for three years (67 production batches) to cover the variability of raw materials and process conditions. NIR calibration models were established respectively, with determination coefficients (r2) of 0.9905, root mean square errors of prediction (RMSEP) of 0.004 for the relative density in the range of 1.042-1.184; r2 of 0.9870, RMSEP of 1.1% for the moisture content in the range of 50.8-83.0%; r2 of 0.9870, RMSEP of 0.461 mg/g for danshensu in the range of 2.563-8.869 mg/g. Then, all the method validation parameters (accuracy, precision, range, specificity, linearity, robustness, detection and quantitation limits) were discussed according to the characteristics of the commercial production of BDPs. The NIR method development and validation strategy proposed may also be applied in the future for the commercial production of other BDPs.

Monitoring batch-to-batch reproducibility of liquid-liquid extraction process using in-line near-infrared spectroscopy combined with multivariate analysis.

Traditional Chinese medicine (TCM) products are usually manufactured through batch processes. To improve batch-to-batch reproducibility, the feasible approaches for real-time monitoring of batch evolution need to be developed. In-line near-infrared (NIR) spectroscopy combined with multivariate data analysis as an efficient process analytical technology (PAT) tool, is presented in this study for real-time batch process monitoring. Liquid-liquid extraction is a widely used purification technology in the TCM manufacture, and selected as the example to demonstrate the effectiveness of this PAT tool. Multi-way partial least squares (MPLS) model was developed based on in-line measured NIR spectral data of ten normal operation condition (NOC) batches. Three kinds of multivariate control charts (scores, Hotelling T(2) and DModX) were used to monitor the evolution of six test batches with artificial batch variations, including the change of starting material quality attributes and abnormal operation conditions. The approach was found very effective for real-time monitoring of process deviations from NOC batches. It is an alternative promising tool for monitoring batch reproducibility of the unit operations during the manufacture of TCM.

Shifting from population-wide to personalized cancer prognosis with microarrays.

The era of personalized medicine for cancer therapeutics has taken an important step forward in making accurate prognoses for individual patients with the adoption of high-throughput microarray technology. However, microarray technology in cancer diagnosis or prognosis has been primarily used for the statistical evaluation of patient populations, and thus excludes inter-individual variability and patient-specific predictions. Here we propose a metric called clinical confidence that serves as a measure of prognostic reliability to facilitate the shift from population-wide to personalized cancer prognosis using microarray-based predictive models. The performance of sample-based models predicted with different clinical confidences was evaluated and compared systematically using three large clinical datasets studying the following cancers: breast cancer, multiple myeloma, and neuroblastoma. Survival curves for patients, with different confidences, were also delineated. The results show that the clinical confidence metric separates patients with different prediction accuracies and survival times. Samples with high clinical confidence were likely to have accurate prognoses from predictive models. Moreover, patients with high clinical confidence would be expected to live for a notably longer or shorter time if their prognosis was good or grim based on the models, respectively. We conclude that clinical confidence could serve as a beneficial metric for personalized cancer prognosis prediction utilizing microarrays. Ascribing a confidence level to prognosis with the clinical confidence metric provides the clinician an objective, personalized basis for decisions, such as choosing the severity of the treatment.