Rapid authentication and differentiation of herbal medicine using ¹H NMR fingerprints coupled with chemometrics

The public concern and increasing awareness of herbal medicines calls for methods available to assure quality and authenticity. In the present study, a novel strategy combing ¹H nuclear magnetic resonance (¹H NMR) spectral analyses coupled with chemometrics was developed and validated to allow comprehensive analysis and rapid authentication of herbal medicines. Polygoni Multiflori Radix, a widely used herbal medicine and a kind of popular functional food, was taken as an example. Characteristic profiling of ¹H NMR fingerprints was achieved by genetic algorithm and a counter-propagation artificial neural network. Multivariate classification methods were employed to evaluate and validate the performance of the developed characteristic profiling of ¹H NMR fingerprints. The results well showed that the proposed method improved the predictive ability of the PLS-DA, and presented equivalent performance to the original spectrum in other classification methods. In conclusion, the developed method is an accurate, sensitive and rapid authentication of Polygoni Multiflori Radix, which could be a useful tool for rapid authentication and quality control of herbal medicines.