基于迁移学习与TrAdaBoost-SVM方法的XRF中药重金属超标研究

doi:10.3964/j.issn.1000-0593(2023)09-2729-05

摘要
参考文献
相关文章 (15)

全文: PDF (2371 KB)
输出: BibTeX | EndNote (RIS)

摘要：中药材重金属超标问题日趋严重，将阻碍中药产业的未来高质量发展，因此研究高效、准确、便捷的超标鉴定方法对于了解中药的安全性具有重要意义。X射线荧光（XRF）光谱分析具有无损检测、快速准确、样品制备方便等优势，在元素分析领域获得广泛应用。由于中药材重金属超标阈值低（如中国药典2020年版规定铅超标为5 mg·kg^-1），中药的种类繁多，基体复杂，国家标准样本匮乏，常规的分类算法难以准确鉴定超标问题。将迁移学习与多分类支持向量机（TrAdaBoost-SVM）方法结合，以金银花为例，采用与金银花相似的国家土壤标准样品的光谱特征信息用于数据增强，将土壤标准样品和少量中药样本混合建立迁移学习和支持向量机分类模型。通过实验验证，迁移学习和TrAdaBoost-SVM结合的分类优化方法，与传统SVM、AdaBoost分类算法相比，鉴定重金属元素铅（Pb）的超标准确率有显著提高。通过测试数据集的预测验证，TrAdaBoost-SVM模型的预测准确率为96.7%，高于传统SVM及AdaBoost分类模型的准确率。所提出的迁移学习与TrAdaBoost-SVM结合的方法，可在小样本条件下建立分类模型，并对中药的重金属超标准确预测，具有一定的理论意义和应用价值。

关键词：X荧光光谱分析技术；迁移学习；支持向量机；中药重金属超标分类

Abstract：The problem of heavy metals exceeding the standard in Chinese medicinal materials is becoming increasingly serious, which will hinder the high-quality development of the Chinese medicine industry in the future. Therefore, research on efficient, accurate and convenient methods for the identification of excessive heavy metals is of great value for understanding the safety of traditional Chinese medicine. X-ray fluorescence spectrometry (XRF) instruments have the advantages of non-destructive testing, fast and accurate, and convenient sample preparation, and are widely used in elemental analysis. Due to the low threshold of heavy metals in traditional Chinese medicinal materials (for example, the 2020 edition of the Chinese Pharmacopoeia stipulates that the lead exceeds the standard at 5 mg·kg^-1), there are many types of traditional Chinese medicines, complex matrices, and lack of national standard samples. Conventional classification algorithms are difficult to identify excessive problems accurately. This paper combines transfer learning with a multi-class support vector machine (TrAdaBoost SVM) method. The spectral feature information of national soil standard samples similar to honeysuckle is used for data enhancement, and the standard soil sample and a small amount of traditional Chinese medicine samples are mixed with establish Transfer learning and support vector machine classification models. Through the experimental verification, the classification optimization method combining transfer learning and TrAdaBoost-SVM, compared with the traditional SVM and AdaBoost classification algorithm, the accuracy rate of identifying the heavy metal element lead (Pb) exceeding the standard has been significantly improved. Through the prediction verification of the test dataset, the prediction accuracy of the TrAdaBoost-SVM model is 96.7%, which is higher than that of the traditional SVM and AdaBoost classification models. The method of combining transfer learning and TrAdaBoost-SVM proposed in this paper can establish a classification model under the condition of small samples and can accurately predict the excess of heavy metals in traditional Chinese medicine, which has certain theoretical significance and application value.

Key words：X-ray fluorescence spectroscopy analysis technology; Migration learning; Support vector machine; Heavy metals in traditional Chinese medicine classification

收稿日期: 2022-04-23 修订日期: 2022-08-31

中图分类号:

O434.13

基金资助: 国家自然科学基金项目(62075028)资助

通讯作者: 李福生 E-mail: lifusheng@uestc.edu.cn

作者简介: 马骞，女，1996年生，电子科技大学自动化工程学院硕士研究生 e-mail: 17366257929@163.com

引用本文:

马骞，杨婉琪，李福生，程惠珠，赵彦春. 基于迁移学习与TrAdaBoost-SVM方法的XRF中药重金属超标研究[J]. 光谱学与光谱分析, 2023, 43(09): 2729-2733.
MA Qian, YANG Wan-qi, LI Fu-sheng, CHENG Hui-zhu, ZHAO Yan-chun. Research on Classification of Heavy Metal Pb in Honeysuckle Based on XRF and Transfer Learning. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2023, 43(09): 2729-2733.

链接本文:

https://www.gpxygpfx.com/CN/10.3964/j.issn.1000-0593(2023)09-2729-05 或 https://www.gpxygpfx.com/CN/Y2023/V43/I09/2729

[1] WANG Hong-bo（王红波）. Digest of the Latest Medical Information in the World（世界最新医学信息文摘），2017，(68): 2.
[2] GUO Lan-ping, ZHOU Li, WANG Shen, et al（郭兰萍, 周利, 王升, 等）. Science and Technology Herald（科技导报）, 2017, 35(11): 8.
[3] YAN Hong-yuan, GONG Wen-ling, LIU Yin, et al（颜鸿远, 龚文玲, 刘引, 等）. Chinese Journal of Experimental Prescriptions（中国实验方剂学杂志）, 2021, 27(12): 10.
[4] DONG Shun-fu, HAN Li-qin, ZHAO Wen-xiu, et al（董顺福, 韩丽琴, 赵文秀, 等）. Chinese Journal of Spectroscopy Laboratory（光谱实验室）, 2010, 27(4): 1346.
[5] ZHOU Bing-wen, ZHU Li-li, ZHU Lin, et al（周炳文, 朱丽丽, 朱林, 等）. Journal of Instrumental Analysis（分析测试学报）, 2021, 40(1): 106.
[6] XU Fei, MENG Sha, WU Qi-nan, et al（徐飞, 孟沙, 吴启南, 等）. Journal of Nanjing University of Chinese Medicine（南京中医药大学学报）, 2018, 34(6): 4.
[7] ZHANG Xiao-ping, ZHU Lai-long（张小平, 朱来龙）. Gansu Normal University Journal（甘肃高师学报）, 2011, 16(5): 20.
[8] LIU Wan-jun, LI Tian-hui, QU Hai-cheng（刘万军, 李天慧, 曲海成）. Remote Sensing for Land & Resources（国土资源遥感）, 2018, 30(4): 41.
[9] XU Shun-gui, LIU Chun-guang（许顺贵, 刘春光）. Chinese Journal of Pharmacovigilance（中国药物警戒）, 2015, 12(10): 4.
[10] SUN Tao（孙桃）. Studies of Trace Elements and Health（微量元素与健康研究）, 2012, 29(4): 2.
[11] CAI Hui-xia, SONG Ya-ling, YANG Meng-cong, et al（蔡慧侠, 宋亚玲, 杨梦聪, 等）. Central South Pharmacy（中南药学）, 2021, 19(2): 282.
[12] Gardner R P，Li F S. X-Ray Spectrometry, 2011, 40(6): 405.
[13] Li F S, Yang W Q, Ma Q, et al. Measurement Science & Technology，2021，32(10)：105501.
[14] Ghidotti M, Papoci S, Dumitrascu C, et al. Talanta Open, 2021, 3: 100040.
[15] XIE Ren-quan, LI Wei, WANG Xian-shu, et al（谢仁权, 李玮, 王贤书, 等）. Journal of Anhui Agricultural Sciences（安徽农业科学）, 2019, 47(9): 4.
[16] FANG Ping, ZOU Wen, SHI Xian-xiao, et al（方萍, 邹雯, 史先肖, 等）. Chinese Journal of Pharmaceutical Analysis（药物分析杂志）, 2017, 37(7): 6.
[17] Lei M, Chen L, Huang B S, et al. Applied Spectroscopy, 2017, 1(1): 2427.
[18] HAN Xiao-li, ZHANG Xiao-bo, GUO Lan-ping, et al（韩小丽, 张小波, 郭兰萍, 等）. China Journal of Chinese Materia Medica（中国中药杂志）, 2008, 33(18): 2041.
[19] TAN Lei, LÜ Hao, ZHAN Yan, et al（谭镭, 吕昊, 詹雁, 等）. China Test（中国测试）, 2009, 35(6): 78.
[20] YUAN De-qing, GAO Peng-jin, RUAN Yi-nan, et al（袁得清, 高鹏锦, 阮毅男, 等）. Journal of Leshan Normal University（乐山师范学院学报）, 2020, 35(4): 33.