EUVIMEDEUVIMED entdecken
Crossref

Three-dimensional fluorescence spectroscopy recognition and component analysis based on machine learning

Zhuohang Wang, Lanying Guo, Zhe Li, Yifan Liu, Kun Cheng, Wei Yang, Xiaohui Wang

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy · 2026-06 · 354 · 127633

DOI: 10.1016/j.saa.2026.127633 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2026-06. Verfasst von Zhuohang Wang und 6 weiteren Autor:innen. Beteiligte Einrichtung: Beijing Key Laboratory of Work Safety Intelligent Monitoring, School of Electronic Engineering, Beijing University of Posts and Telecommunications, B…. Die Arbeit wurde bisher 6-mal zitiert (Crossref) und nennt 30 Literaturangaben. Gefördert von: Fundamental Research Funds for the Central Universities, National Natural Science Foundation of China.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Three-dimensional excitation-emission matrix (EEM) fluorescence spectroscopy has received extensive attention in the identification of fluorescent substances due to its advantages of low cost, convenient operation and high sensitivity. However, the contamination by impurities with similar spectra can seriously affect the accuracy and reliability of the identification. In this study, we present an effective approach integrating machine learning with three-dimensional EEM spectra to achieve contamination detection and component analysis of samples. The EEM fluorescence spectral data were collected in a simulated contaminated environment using a fluorescence spectrophotometer. The detection and identification effects of various contaminants were assessed using multiple algorithms including K-Nearest Neighbors (KNN), Random Forest (RF) and Convolutional Neural Network (CNN). The analysis of various fluorescent components in contaminated samples was achieved by using PARAFAC-derived spectral profiles as reference targets to train an improved U-shaped Network (U-Net) architecture, named Shared Encoder U-Net (SE-UNet). The analysis performance was evaluated using Root Mean Squared Error (RMSE), Normalized Root Mean Squared Error (NRMSE), cosine similarity and a Weighted Mean Absolute Error (WMAE) that emphasizes peak regions. In relatively simple binary-component scenarios, all deep learning models achieved comparable accuracy. As the mixture composition became more complex, the optimized SE-UNet exhibited more stable performance and stronger generalization than models based on Baseline Convolutional Neural Networks (CNN) or the Visual Geometry Group (VGG) architecture, while enabling fast single-sample inference compared with the iterative batch Parallel Factor Analysis (PARAFAC) method. This framework provides a practical and scalable solution for contamination analysis in laboratory and environmental monitoring scenarios.

Schlagwörter

Stichwörter der Autor:innen

Component analysisMachine learningPARAFACSample contaminationThree-dimensional excitation-emission matrix

Autor:innen und Institutionen

  • Beijing Key Laboratory of Work Safety Intelligent Monitoring, School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China.
  • School of Medical Imageology, Wannan Medical College, Wuhu 241002, China.
  • Beijing Key Laboratory of Work Safety Intelligent Monitoring, School of Electronic Engineering, Beijing University of Posts and Telecommunications, Beijing 100876, China. Electronic address: wangxiaohui@bupt.edu.cn.

Förderung

  • Fundamental Research Funds for the Central Universities
  • National Natural Science Foundation of China

Zitieren

APA

Wang, Z., Guo, L., Li, Z., Liu, Y., Cheng, K., Yang, W., & Wang, X. (2026). Three-dimensional fluorescence spectroscopy recognition and component analysis based on machine learning. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 354, 127633. https://doi.org/10.1016/j.saa.2026.127633

Vancouver

Wang Z, Guo L, Li Z, Liu Y, Cheng K, Yang W, et al. Three-dimensional fluorescence spectroscopy recognition and component analysis based on machine learning. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. 2026;354:127633. doi:10.1016/j.saa.2026.127633

BibTeX
@article{Wang2026,
  author = {Wang, Zhuohang and Guo, Lanying and Li, Zhe and Liu, Yifan and Cheng, Kun and Yang, Wei and Wang, Xiaohui},
  title = {Three-dimensional fluorescence spectroscopy recognition and component analysis based on machine learning},
  journal = {Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy},
  year = {2026},
  volume = {354},
  pages = {127633},
  doi = {10.1016/j.saa.2026.127633}
}
Kostenlos starten →