[Classification of Chinese traditional medicine, moutan (Paeonia suffruticosa Andr.) cortex by means of pyrolysis-high resolution gas chromatography/pattern recognition].

Thirty-eight samples of Moutan (Paeonia suffruticosa Andr.) Cortex obtained from three different regions (Southwest, East and Middle China) and two samples of unknown region were subjected to analysis with pyrolysis-high resolution gas chromatography (Py-HRGC). Each sample was thus characterized by the peak area of 41 peaks in each Py-HRGC profile. Discriminant analysis (DA), PRIMA and SIMCA pattern recognition were used to recognize the 40 x 41 data matrix. These data analysis gave satisfactory results (DA, 100% correct; PRIMA, 100% correct, 92.2% unique; SIMCA, 92.2% correct, 79% unique). The correct classification of Moutan Cortex for the unknown territory was obtained by three pattern recognition methods. The results showed that Py-HRGC/pattern recognition technique might be a potential tool for the identification of Chinese traditional medicine.