用于降糖类药物真伪快速检测的3种拉曼光谱处理方法的比较

Algorithms of Raman spectrium analysis in fast identification of counterfeit hypoglycemic agents

目的比较3种用于药品真伪快速检测的拉曼光谱处理方法的适用性。方法采用必达泰克便携式拉曼光谱仪测定6类降糖类药物对照品和相应的市售药品与12种常用制剂辅料的拉曼光谱,采用上海技术物理研究所专利改进方法(MSFDSS)、局部直线筛选法(LSLS)和改进的局部直线筛选法(MLSLS)等3种处理方法对光谱进行处理。结果 MLSLS法处理结果的灵敏度、专属性、阴性预报性和检出效率分别为96.77%、97.20%、99.29%和96.02%,总体结果在3种处理方法中最佳。结论发源于红外光谱的LSLS方法经过改进后,可以用于药品真伪的拉曼光谱法快速检测。

Objective To investigate the applicability of portable Raman spectroscopy for fast identification of counterfeit hypoglycemic agents. Methods The Raman spectra of 6 hypoglycemic agents, their corresponding reference standards and 12 common pharmaceutical excipients were detected by BW＆Tek i-Raman portable spectrometer. The spectra were analyzed by three algorithms, namely, the minimized summation of first derivative of subtraction spectra algorithm, local straight-line screening （LSLS） algorithm, and modified LSLS （MLSLS） algorithm. Results Among the three algorithms, MLSLS algorithm showed the best performance, with the total sensitivity, specificity, accuracy and efficiency being 96.77%, 97.20%, 99.29%, and 96.02%, respectively. Conclusion The MLSLS algorithm, which is based on infrared spectroscopy, can be adapted to identify suspected counterfeit drugs based on portable Raman spectroscopy.