Rapid quality evaluation of Saposhnikoviae Radix powders by NIRS combined with multivariate intelligent algorithms

Saposhnikoviae Radix is a renowned herb, with a long history and both medicinal and edible value, but it's quality is affected by various factors such as geographical sources, harvest years, and wild or cultivated. The total content of prim-O-glucosylcimifugin and 5-O-methylvisammioside is the...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Microchemical journal Jg. 215; S. 114559
Hauptverfasser: Xue, Zhong, Bai, Jie, Luo, Yunxiao, Zhao, Qianzhou, Cao, Chunqi, Huang, Yun
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier B.V 01.08.2025
Schlagworte:
Multivariate intelligent algorithms
NIRS
Quality evaluation
Saposhnikoviae Radix
NIRS
Saposhnikoviae Radix
Quality evaluation
Multivariate intelligent algorithms
ISSN:0026-265X
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Saposhnikoviae Radix is a renowned herb, with a long history and both medicinal and edible value, but it's quality is affected by various factors such as geographical sources, harvest years, and wild or cultivated. The total content of prim-O-glucosylcimifugin and 5-O-methylvisammioside is the important quality index of Saposhnikoviae Radix. For conventional liquid-phase analysis methods, we need to analyze the content of prim-O-glucosylcimifugin and 5-O-methylvisammioside separately and then calculate the total amount. In this paper, we employed the near infrared spectroscopy (NIRS) for rapid quality evaluation of Saposhnikoviae Radix by directly quantifying the total amount of prim-O-glucosylcimifugin and 5-O-methylvisammioside, instead of quantifying the prim-O-glucosylcimifugin and 5-O-methylvisammioside separately. The PLS regression quantitative model was used for correlating the total amount of prim-O-glucosylcimifugin and 5-O-methylvisammioside and NIR spectral data. Multivariate intelligent algorithms, including 15 spectral preprocessing methods and 12 variables selection methods, were employed to improve the performance of PLS quantitative model. The results displayed that the PLS model after joint optimization strategy of MSC + 1D + CARS, showed a significant improvement in quantitative and predicting performance, and the developed NIR method can directly, quickly, and accurately evaluate the quality of Saposhnikoviae Radix successfully. [Display omitted] •A rapid, non-destructive and reliable NIR quantitative method is developed.•The quality of Saposhnikoviae Radix powders is assessed by NIR successfully.•Multivariate intelligent algorithms are used to improve the performance of PLS model.•The established NIR method after MSC + 1D + CARS has excellent predictive performance.
ISSN:0026-265X
DOI:10.1016/j.microc.2025.114559