浏览全部资源
扫码关注微信
1.云南中医药大学 中药学院,云南 昆明 650500
2.云南省农业科学院 药用植物研究所, 云南 昆明 650223
王元忠,博士,副研究员,研究方向:中药资源开发与利用,E-mail:boletus@126.com
纸质出版日期:2024-11-15,
收稿日期:2024-05-17,
修回日期:2024-07-01,
移动端阅览
胡晓燕,王元忠.基于傅里叶变换中红外光谱的不同维度光谱图像结合残差神经网络鉴别黄精属物种[J].分析测试学报,2024,43(11):1709-1724.
HU Xiao-yan,WANG Yuan-zhong.Identification of Species in the Polygonatum Genus Based on Different Dimensional Spectral Images Combined with Residual Neural Networks in Fourier Transform Infrared Spectroscopy[J].Journal of Instrumental Analysis,2024,43(11):1709-1724.
胡晓燕,王元忠.基于傅里叶变换中红外光谱的不同维度光谱图像结合残差神经网络鉴别黄精属物种[J].分析测试学报,2024,43(11):1709-1724. DOI: 10.12452/j.fxcsxb.24051750.
HU Xiao-yan,WANG Yuan-zhong.Identification of Species in the Polygonatum Genus Based on Different Dimensional Spectral Images Combined with Residual Neural Networks in Fourier Transform Infrared Spectroscopy[J].Journal of Instrumental Analysis,2024,43(11):1709-1724. DOI: 10.12452/j.fxcsxb.24051750.
将不同维度光谱图像的概念首次应用于物种鉴别,建立了快速准确的黄精属鉴别方法。采集6种黄精属共计563批样品,基于傅里叶变换中红外光谱(FT-MIR)的一阶导数(1st)、二阶导数(2nd)、乘法散射校正(MSC)、标准正态变量变换(SNV)和Savitzky-Golay(SG)5种预处理方法,构建了决策树(DT)、随机森林(RF)和支持向量机(SVM)3种机器学习算法。同时构建了深度学习残差神经网络(ResNet)模型,绘制了不同维度的光谱图像,包括一维MIR,同步、异步和综合二维相关光谱、三维相关光谱、三维相关光谱投影图像的10个数据集,并将其与ResNet模型相结合进行分类。结果表明,不同预处理方法对模型结果的影响不同,MSC预处理方法可显著提高DT、RF和SVM 3种算法的准确率。基于同步二维相关光谱数据集的ResNet算法建模效果最好,准确率达到100%,损失值较小,不需要复杂的预处理,时间成本低,可以准确鉴别黄精属物种,为食品、中草药等其他领域的鉴别提供了参考。
This study proposed the concept of spectral images with different dimensions,which was first applied to the identification of species in the
Polygonatum
genus,and established a fast and accurate identification method. A total of 563 batches of samples from 6 species of
Polygonatum
were collected. Five preprocessing methods were used based on Fourier transform mid infrared spectroscopy(FT-MIR),including first derivative(1st),second derivative(2nd),multiplicative scattering correction(MSC),standard normal variable(SNV) and Savitzky-Golay(SG). Decision trees(DT),random forests(RF) and support vector machines(SVM) were constructed. In addition,to avoid complex preprocessing,a deep learning residual convolutional neural network(ResNet) model was constructed to draw spectral images of different dimensions,including 10 datasets of one-dimensional MIR,synchronous,asynchronous,and integrative two-dimensional correlated spectra,three-dimensional correlated spectra,and three-dimensional correlated spectral projection images,and they were combined with the ResNet model for classification. The results showed that different preprocessing methods have different impacts on the model results,and the MSC preprocessing method significantly improved the accuracy of the DT,RF,and SVM algorithms. The ResNet algorithm based on synchro
nous two-dimensional correlated spectral dataset had the best modeling effect,with an accuracy of 100%,small loss value,no need for complex preprocessing,low time cost,and could accurately identify species of
Polygonatum
genus,providing reference for identification in other fields such as food and traditional Chinese medicine.
黄精属物种傅里叶变换中红外光谱不同维度光谱图像机器学习算法残差神经网络
species of Polygonatum genusFourier transform mid infrared spectroscopyspectral images of different dimensionsmachine learning algorithmsresidual neural networks
Zhang J,Wang Y Z,Yang W Z,Yang M Q,Zhang J Y. China J. Chin. Mater. Med. (张娇,王元忠,杨维泽,杨美权,张金渝. 中国中药杂志),2019,44(10):1989-2008.
Li R S,Tao A E,Yang R M,Fan M,Zhang X C,Du Z F,Shang F N,Xia C L,Duan B Z. Biomed. Pharmacother.,2020,131:110687.
Cui X W,Wang S Y,Cao H,Guo H,Li Y J,Xu F X,Zheng M M,Xi X Z,Han C C. Molecules,2018,23(5):1170.
National Pharmacopoeia Commission. Pharmacopoeia of People's Republic of China(Part I). Beijing:China Medical Sci-ence Press(国家药典委员会.中华人民共和国药典(1部).北京:中国医药科技出版社),2020:319.
Dong J E,Zhang J,Zuo Z T,Wang Y Z. Spectrochim. Acta A,2021,249:119211.
Yue J Q,Huang H Y,Wang Y Z. Microchem. J.,2021,160:105731.
Li C H,Luo W,Xu H Y,Zhang Q Xu H,Aguilar Z P,Lai W H,Wei H,Xiong Y H. Food Control,2013,34(2):725-732.
Yang X C,Chen L F,Wang H W,Zang H C,Guo X P,Nie L. J. Instrum. Anal.(杨向春,陈丽芳,王浩伟,臧恒昌,郭学平,聂磊. 分析测试学报),2023,42(8):943-951.
Vodnar D C,Socaciu C,Rotar A M,Stãnilã A. Int. J. Food Sci. Technol.,2010,45(11):2345-2351.
Chen L,Chang L F,Shen M A,Zhang M Y,Li D J,Li Y J. J. Instrum. Anal.(陈露,常龙飞,沈沐傲,张鸣原,李德建,李英骏. 分析测试学报),2024,43(3):489-495.
Wang L G,Long F,Liao W,Liu H. Bioresour. Technol.,2020,298:122495.
Xu Y L,Yang W Z,Wu X W,Wang Y Z,Zhang J Y. Foods,2022,11(22):3568.
Wang Y D,Jia J W,Tan W J,Lei M. J. Instrum. Anal.(王亚栋,贾俊伟,谭韦君,雷萌.分析测试学报),2024,43(4):607-613.
Noda I. Spectrochim. Acta A,2018,197:4-9.
Zhu Y,Xu C H,Huang J,Li G Y,Liu X H,Sun S Q,Wang J H. J. Mol. Struct.,2014,1069:272-279.
Li L,Zhao Y L,Li Z M,Wang Y Z. Vib. Spectrosc.,2022,123:103429.
Jiao Y P,Li Z C,Chen X S,Fei S M. J. Chemom.,2020,34(11):e3306.
Bi Y M,Yuan K L,Xiao W Q,Wu J Z,Shi C Y,Xia J,Chu G H,Zhang G X,Zhou G J. Anal. Chim. Acta,2016,909:30-40.
Noda I. Appl. Spectrosc.,1993,47(9):1329-1336.
Noda I. J. Mol. Struct.,2018,1160:471-478.
Charbuty B,Abdulazeez A. J. Appl. Sci. Technol. Trends,2021,2(1):20-28.
Dong J E,Li J Q,Liu H G,Wang Y Z. Food Res. Int.,2023,167:112679.
Zhang Y Y,Wang Y Z. Food Chem. X,2023,19:100860.
Wang L,Li J Q,Li T,Liu H G,Wang Y Z. ACS Omega,2021,6(30):19665-19674.
Wu X M,Zhang Q Z,Wang Y Z. Spectrochim. Acta A,2018,205:479-488.
Zhang Y Y,Wang Y Z. J. Pharm. Anal.,2023,13(12):1388-1407.
Chen J B,Guo B L,Yan R,Sun S Q,Zhou Q. Spectrochim. Acta A,2017,182:81-86.
Pei Y F,Wu L H,Zhang Q Z,Wang Y Z. Anal. Methods,2019,11(1):113-122.
Qin X J,Zhang L J,Zhang Y,Ni W,Yang X Z,Yu Q,Yan H,An L K,Liu H Y. Bioorg. Chem.,2020,99:103788.
Chen J,Li J Q,Li T,Liu H G,Wang Y Z. Crit. Rev. Anal. Chem.,2021,53(4):852-868.
Dong J E,Zuo Z T,Zhang J,Wang Y Z. Food Control,2021,129:108132.
Qi L M,Liu H G,Li J Q,Li T,Wang Y Z. Sensors,2018,18(1):241.
Noda I. J. Am. Chem. Soc.,2002,111(21):8116-8118.
0
浏览量
27
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构