| In this dissertation,we mainly established qualitative and quantitative models based on near-infrared spectroscopy and color difference meter combined with chemometrics,which were used to distinguish the degree of suitability of processed Rhizoma Atractylodis Macrocephalae(referred as BZ below)and to determine the contents of three atractylenolides,which thus are of great help to establish a fast,simple and reliable method for the quality evaluation of BZ decoction pieces.The main contents and results of this paper are as follows:1.Twelve batches of raw medicinal materials of BZ were collected from different places of origin.According to the Pharmacopoeia of the People's Republic of China,these raw decoction pieces were stir-fried in wheat bran using a stir-frying machine for 3,6,9,12 and 15 minutes,respectively.The resulted 60 samples were categorized into three classes(i.e.,light,moderate and dark)by experienced pharmacists according to their surface color or the crashed powder color.The labelled color data were denoted as Y1 and Y2,respectively.After the acquisition of the color difference data of the BZ powder(denoted as X)by the colorimeter,two qualitative models were constructed by partial least squares discriminant analysis(PLS-DA)to relate X to Y1 and Y2,respectively.The PLS-DA models were evaluated by jackknife tests with calculating parameters such as error rate(ERR),true positive rate(TPR),true negative rate(TNR)and F1 score(F1).The results showed that the X-Y2 PLS-DA model performed better than X-Y1,indicating that the surface color of the processed BZ was not consistent with their inner color,especially when the raw slices have different size and thickness.2.The spectra of BZ was collected by near-infrared spectroscopy,denoted as X'.Qualitative models were also constructed by PLS-DA to relate X'to Y1.Various spectra pretreatment methods like standard normal variate,multiplicative scatter correction,derivation and smoothing,and fourteen feature selection methods were compared and optimized in detail.The results showed that the PLS-DA models with spectra pretreated by1D5S or 1D9S and wavebands selected by InfFS,Relief-F,MutInfFS,Fisher or CFS performed best,yielding 0.00 of ERR,1.00 of TPR,1.00 of TNR,and 1.00 of F1 for all three classes.Thus,the present qualitative models can be successfully used to distinguish the degree of suitability of processed BZ,which provide a more objective method for the fast quality evaluation and control of processed BZ and other decoction pieces.3.A HPLC method for determination of atractylenolides I,II and III was established and validated.After that,the contents of the three atractylenolides in processed BZ samples were determined and denoted as Y3.Quantitative models were constructed by partial least squares regression(PLSR)to relate X or X'to Y3,respectively.The PLSR models were optimized by spectral pretreatment method and feature selection methods such as particle swarm optimization algorithm and genetic algorithm(GA)to improve the performance.Meanwhile,the PLSR models were evaluated with calculating parameters such as coefficients of determination(R~2),root mean square error(RMSE),mean absolute error(MAE),and residual predictive deviation(RPD).The results showed that the X'-Y3 model performed better than the X-Y3 model,especially when X'was with 1D9S spectra pretreatment and GA wavebands selection.R_C~2and R_P~2 were all>0.95,RMSE_C and RMSE_P were all<0.04%,MAE_C and MAE_P were all<0.04%,and RPD were all>5.Thus,the present quantitative models constructed by the near-infrared spectroscopy can be used to accurately determine the contents of the atractylenolides I,II and III in BZ decoction pieces,which are of great help for fast quality evaluation of processed BZ and other decoction pieces. |
PDF Full Text Request