Comprehensive Application of AHP-CRITIC Hybrid Weighting Method， Grey Correlation Analysis and BP-ANN in Optimization of Extraction Process of Qizhi Prescription
10.13422/j.cnki.syfjx.20250667
- VernacularTitle:AHP-CRITIC混合加权法、灰色关联度分析与反向传播人工神经网络在芪志方提取工艺优化中的综合应用
- Author:
Qun LAN
1
;
Yi CHENG
1
;
Zian LI
1
;
Bingyu WU
1
;
Jinyu WANG
1
;
Dewen LIU
1
;
Yan TONG
1
Author Information
1. Institute of Chinese Materia Medica，China Academy of Chinese Medical Sciences，Beijing 100700，China
- Publication Type:Journal Article
- Keywords:
Qizhi prescription;
analytic hierarchy process（AHP）;
criteria importance through intercriteria correlation（CRITIC）;
orthogonal test;
grey relational analysis;
backpropagation artificial neural network（BP-ANN）;
extraction process
- From:
Chinese Journal of Experimental Traditional Medical Formulae
2025;31(8):176-186
- CountryChina
- Language:Chinese
-
Abstract:
ObjectiveBased on analytic hierarchy process（AHP）-criteria importance through intercriteria correlation（CRITIC） hybrid weighting method， grey relational analysis and backpropagation artificial neural network（BP-ANN）， to optimize the water extraction process of Qizhi prescription， so as to provide an experimental basis for optimization of the preparation process of this prescription and the establishment of quality standards. MethodsL9（34） orthogonal test was employed， and the AHP-CRITIC hybrid weighting method was utilized to determine the weight coefficients of the quality fractions of various components， including astragaloside Ⅳ， polygalaxanthone Ⅲ， calycosin-7-O-β-D-glucoside， tenuifolin， and 3，6′-disinapoylsucrose， as well as the dry extract yield. The comprehensive score of each factor level combination in the orthogonal test were calculated as evaluation indicator to select the optimal extraction process parameters. The effects of extraction times， extraction time， and solvent dosage on the aqueous extraction process of the formula were investigated through intuitive analysis， variance analysis， and grey relational analysis. Meanwhile， a BP-ANN model was established to reverse-predict the optimal extraction process parameters of Qizhi prescription， and the optimized process parameters were validated. ResultsThe weight coefficients of the five index components（astragaloside Ⅳ， tenuifolin， calycosin-7-O-β-D-glucoside， polygalaxanthone Ⅲ， and 3，6′-disinapoylsucrose） and dry extract yield were 25.7%， 20.82%， 16.41%， 12.45%， 15.96% and 8.67%， respectively. The optimized extraction process parameters were extracted 3 times with 8， 6， 6 times the amount of water， each time for 1 h. The network prediction results of BP-ANN test samples were consistent with the orthogonal test results， and the mean square error（MSE） of the predicted and measured values of the network was <1%. The water extraction process of Qizhi prescription analyzed and predicted by relevant mathematical models was stable and feasible， which could effectively improve the extraction efficiency of the active ingredients of Astragali Radix and Polygalae Radix， and the average comprehensive score of the validation test was 90.85 with the relative standard deviation（RSD） of 1.55%. ConclusionThis study establishes a water extraction process for compound Qizhi granules， and the optimized extraction process can effectively improve the extraction efficiency of active ingredients， which provides useful references for the optimization of preparation process and the establishment of quality standards for other clinical experience formulas.
