应用近红外光谱法建立伤疖膏制备过程中黄芩苷含量模型

建立一种伤疖膏制备过程提取液中黄芩苷动态含量快速测定的近红外光谱分析方法，近红外透射光谱法扫描得到65组伤疖膏制备过程中提取液的近红外光谱图，以提取液中黄芩苷的HPLC测量值作为对照值，采用偏最小二乘回归算法(PLSR)建立NIR光谱与对照值的校正模型。校正模型主成分数为8，交叉验证均方根差(RMSECV)为0.006 8，相关系数(r)为0.999 1。应用校正模型对预测集的30组样品进行黄芩苷含量预测，所得预测均方根差(RMSEP)为0.009 2，r为0.998 7。结果表明，该方法快速、准确，为复方膏剂制备过程中化学成分快速定量和质量控制提供了方法和依据。

Quantitative Models for Baicalin Content Using NIR Technology for the Study of Shang Jie Plaster

A dynamic prediction model for the content of Baicalin in Shang Jie plasters extract solutions was developed using near-infrared spectroscopy in transmission mode. Sixty five spectra were obtained through near-infrared transmission mode during extracting process. Refering to the content of Baicalin performed by reversed-phase high performance liquid chromatography (HPLC), the calibration model was developed with the application of partial least squares regression algorithm (PLSR). The constructed model was validated by 30 samples; some parameters of the calibration model were optimized by cross-validation. The root mean square error (RMSECV) of Baicalin was 0.006 8 mg·g-1, the correlation coefficient (R) was 0.9991, and the optimal dimension factor was 8; After predicted by test set, the root mean square error (RMSEP) and correlation coefficient (R) of prediction obtained were 0.009 2 mg·g-1 and 0.998 7 respectively. This work demonstrated that NIR spectroscopy combined with PLS could be used for the determination of Baicalin in Shang Jie plasters extract.