Determination of Paeoniflorin in Qizhiweitong Granules Based on Near Infrared Spectroscopy Combined with Chemometrics

目的：研究近红外光谱在一步制粒过程中监测气滞胃痛颗粒芍药苷含量的可行性。方法：利用近红外在线检测系统，非接触式探头对制粒过程进行全程近红外光谱扫描，对采集到的光谱进行预处理和波段选择，并结合偏最小二乘法(partial least squares, PLS)建立芍药苷含量快速无损检测方法。结果：所建立的模型的决定系数R为0.9172，交叉验证均方根差值为0.535，验证集样品预测统计分析得出预测值与真实值之间无显著差异(P > 0.05)。结论：所建立的模型准确度高，适用于气滞胃痛一步制粒过程中芍药苷含量的实时预测。 Objective: To study the feasibility of paeoniflorin content in Qizhiweitong granules by near infrared spectroscopy in one-step granulation process. Methods: Using near infrared online detection system, the non-contact probe was used to scan the whole process of granulation by near-infrared spectroscopy. The spectra of the samples were pretreated and the bands were selected. A fast and nondestructive method for the paeoniflorin content was established by using partial least squares (PLS). Results: The determination coefficient R of the established model was 0.9172, and the root mean square difference of cross-validation was 0.535. The samples of the validation set were predicted and statistically analyzed, and there was no significant difference between the predicted value and the true value (P > 0.05). Conclusion: The established model has high accuracy and is suitable for real-time prediction of paeoniflorin content in one-step granulation process of Qizhi Weitong granules.