浏览全部资源
扫码关注微信
1.暨南大学 物理与光电工程学院,广东 广州 510632
2.广西医科大学附属肿瘤医院,广西 南宁 530021
3.北京理工大学 重庆微电子研究院,重庆 400000
潘涛,博士,教授、博士生导师,研究方向:光谱技术、光谱分析与化学计量学,E-mail:466945939@qq.com
收稿日期:2025-02-27,
修回日期:2025-04-03,
录用日期:2025-04-07,
纸质出版日期:2025-06-15
移动端阅览
高乔基,吴振邦,徐茜,陈敏,刘文轩,曹诚诚,廖敬龙,欧超,潘涛.血清HBsAg感染的Vis-NIR光谱模式识别研究[J].分析测试学报,2025,44(06):1016-1023.
GAO Qiao-ji,WU Zhen-bang,XU Xi,CHEN Min,LIU Wen-xuan,CAO Cheng-cheng,LIAO Jing-long,OU Chao,PAN Tao.Study on Vis-NIR Spectral Pattern Recognition of Serum HBsAg Infection[J].Journal of Instrumental Analysis,2025,44(06):1016-1023.
高乔基,吴振邦,徐茜,陈敏,刘文轩,曹诚诚,廖敬龙,欧超,潘涛.血清HBsAg感染的Vis-NIR光谱模式识别研究[J].分析测试学报,2025,44(06):1016-1023. DOI: 10.12452/j.fxcsxb.250227129.
GAO Qiao-ji,WU Zhen-bang,XU Xi,CHEN Min,LIU Wen-xuan,CAO Cheng-cheng,LIAO Jing-long,OU Chao,PAN Tao.Study on Vis-NIR Spectral Pattern Recognition of Serum HBsAg Infection[J].Journal of Instrumental Analysis,2025,44(06):1016-1023. DOI: 10.12452/j.fxcsxb.250227129.
乙肝表面抗原(HBsAg)是乙肝病毒感染的重要标志物。该文建立了血清HBsAg感染的无试剂可见-近红外(Vis-NIR)光谱模式识别新方法。收集到临床血清样品1 243例(HBsAg阳性601、阴性642),采用训练-预测-检验实验设计,搭建了基于多尺度卷积、压缩-激励网络(SE Net)注意力机制和多尺度膨胀卷积的新型卷积神经网络(CNN)集成算法,连同经典的偏最小二乘-判别分析(PLS-DA)和普通浅层CNN算法,被用于建立HBsAg阳性和阴性血清的Vis-NIR光谱判别模型。该研究采用标准正态变量(SNV)变换进行光谱预处理。基于近红外区(780~1 118 nm)经SNV处理的光谱的PLS-DA模型和新型CNN模型取得更优的建模效果,新型CNN模型的灵敏度(SEN)达到99.3%,漏诊率(FNR)达到0.7%。结果表明,采用Vis-NIR光谱精准判别HBsAg阳性和阴性血清具有可行性,提出的新型深度学习算法可望应用于其他光谱分析领域。
Hepatitis B surface antigen(HBsAg) is an important marker of hepatitis B virus infection. In this article,a novel method of reagent-free visible-near-infrared(Vis-NIR) spectral pattern recognition for serum HBsAg infection was studied. A total of 1 243 clinical serum samples(HBsAg positive 601 and negative 642) were collected,and a training-prediction-validation experimental design was used. A novel CNN integrated algorithm based on multi-cale convolution,SE Net attention mechanism and multi-scale dilated convolutions was constructed,which together with the classic partial least squares-discriminant analysis(PLS-DA) and the ordinary shallow CNN algorithm,were used to establish the Vis-NIR spectral discrimination model for HBsAg positive and negative serums. The standard normal variable(SNV) transform was used for spectral preprocessing. The PLS-DA and new types of CNN models based on the SNV spectra of near-infrared region(780-1 118nm) achieved significantly better modeling results,and the sensitivity(SEN) of the new CNN model reached a significantly higher 99.3%,and the false negative rate(FNR) reached a significantly lower 0.7%. The results show the feasibility of using serum Vis-NIR spectra to accurately identify HBsAg infection,and the proposed new types of deep learning algorithm is also promising for application in other spectral analysis fields.
Bec' K B , Grabska J , Huck C W . Molecules , 2020 , 25 ( 12 ): 2948 .
Fonseca A C S , Pereira J F Q , Honorato R S , Bro R , Pimentel M F . Spectrochim. Acta A , 2022 , 267 : 120533 .
Yao L J , Xu W Q , Pan T , Chen J M . J. Innovative Opt. Health Sci. , 2018 , 11 ( 2 ): 1850005 .
Yuan L , Zhang J , Xu J H , Yao L J , Wang D W , Pan T . Sense the Real Change:Proceedings of the 20th International Conference on Near Infrared Spectroscopy . Singapore : Springer , 2022 : 87 - 192 .
Wang Q , Wang T , Cao L , Mu A , Fu S , Wang P , Gao Y , Ji W X , Liu Z Y , Du Z Q , Guddat L W , Zhang W C , Li S , Li X M , Lou Z Y , Wang X X , Hu Z Y , Rao Z H . Science , 2024 , 385 ( 6714 ): 1217 – 1224 .
Zeng J , Guo Y , Han Y Q , Li Z M , Yang Z X , Chai Q Q , Wang W , Zhang Y Y , Fu C L . Molecules , 2021 , 26 ( 3 ): 749 .
Gan F , Luo J . Chemom. Intell. Lab. , 2023 , 232 : 104710 .
Huang H H , Luo J F , Gan F , Hopke P K . Appl. Sci. , 2023 , 13 ( 14 ): 8494 .
Zhu Y K , Lu C H , Zhang Y J , Jiang W W . Spectrosc. Spectral Anal. (朱御康,鲁昌华,张玉钧,蒋薇薇. 光谱学与光谱分析), 2024 , 9 : 2607 - 2612 .
Deng T M , Liu X H , Wang L . Electronics-Switz. , 2022 , 11 ( 17 ): 2709 .
Hu J , Shen L , Albanie S , Sun G , Wu E . IEEE Trans. Pattern Anal. Mach. Intell. , 2020 , 42 ( 8 ): 2011 – 2023 .
He K M , Zhang X Y , Ren S Q , Sun J . 2016 IEEE Conference on Computer Vision and Pattern Recognition(CVPR) , Las Vegas,USA , 2016 : 770 - 778 .
Liu H T , Hong X P , Xi J , Menne S , Hu J M , Wang J C Y . Sci. Adv. , 2022 , 8 ( 31 ): eabo4184 .
Gu J X , Wang Z H , Kuen J , Ma L Y , Shahroudy A , Shuai B , Liu T , Wang X X , Wang G , Cai J F , Chen T H . Pattern. Recognit. , 2018 , 77 : 354 - 377 .
Barnes R J , Dhanoa M S , Lister S J . Appl. Spectrosc. , 1989 , 43 ( 5 ): 772 – 777 .
Dhanoa M S , Lister S J , Sanderson R , Barnes R J . J. Near Infrared Spectrosc. , 1994 , 2 ( 1 ): 43 – 47 .
Huang Y , Liu H , Lu X , Yao L , Chen J , Pan T . Spectrochim. Acta A , 2024 , 317 : 124394 .
Puertas G , Cazón P , Vázquez M . LWT Technol. , 2023 , 174 : 114429 .
Chen J M , Peng L J , Han Y , Yao L J , Zhang J , Pan T . Spectrochim. Acta A , 2018 , 193 : 499 - 506 .
0
浏览量
17
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构