Origin traceability of Angelica dahurica based on multidimensional bionic sensory, near infrared spectroscopy combined with multi-source information fusion research on rapid identification method of index components.

As a bulk medicinal material and food spice, the quality of Angelicae Dahuricae Radix depends strongly on its producing area. Traditional origin identification relies on subjective judgment, while modern analytical techniques are costly and complex. Therefore, we aimed to develop a rapid, accurate method to identify the origin of Angelicae Dahuricae Radix and to predict the contents of its index components using multi-source information fusion. We integrated multi-dimensional bionic sensory (BS) technologies - electronic eye (EE), electronic nose (EN), electronic tongue (ET), and near-infrared spectroscopy (NIR) - to capture visual, olfactory, gustatory, and spectral information from 81 batches of Angelica dahurica samples collected from four major producing areas in China (Sichuan, Anhui, Henan, Hebei). We fused these multi-level data using multi-source information fusion (MIF). For qualitative origin discrimination, we built models on single-source and fused data using partial least squares discriminant analysis (PLS-DA), least squares support vector machine (LS-SVM), and convolutional neural network (CNN). For quantitative prediction of index components (Bergapten, Oxypeucedanin, Imperatorin, Phellopterin, Isoimperatorin), we constructed back-propagation neural network (BPNN) models. The results demonstrated that, within the MIF framework, the qualitative discriminant model achieved a 100% positive classification rate, markedly outperforming models based on single information sources. Among the quantitative prediction models, the ET-based Phellopterin model exhibited the highest predictive accuracy (Rp(Nan et al., 20262) = 0.9026). Incorporating MIF further enhanced the predictive performance of the Bergapten, Oxypeucedanin, and Imperatorin models, increasing Rp(Nan et al., 20262) values by 28.15%, 10.20%, and 28.63%, respectively. In conclusion, this study established a novel approach for origin traceability and quality evaluation of Angelica dahurica by integrating multi-dimensional BS data with NIR. The proposed method offers rapid and high-precision analysis, providing a robust technical framework and valuable reference for developing quality control and evaluation systems for Chinese medicinal materials and their decoction pieces.