Construction of a universal model for non-invasive identification of cephalosporins for injection using near-infrared diffuse reflectance spectroscopy

A universal classification model has been developed for identification of 26 different products of cephalosporins for injection using near-infrared (NIR) spectroscopy. A total of 324 batches of the 26 products, from 166 manufacturers in China, were used in the study. The classification model was a principle component analysis (PCA)-based method consisting of a primary identification library with four sub-libraries. The accuracy and specificity of the model were tested and were both found to be approximately 95%. The transfer of the classification model between different instruments of the same brand and the same instrument model has been investigated in this study. This study has shown that it is feasible to build a universal classification model to quickly screen for counterfeit drugs in the open market and distribution channels.