EUVIMEDEUVIMED entdecken
Crossref

A Simulation-Driven Deep Learning Approach for Artifact-Resistant Raman Spectral Unmixing

Yunze Li iD, Jiaqi Hu, Tianyu Lai, Lilian Wang, Perry Ping Shum, Gina Jinna Chen iD

Analytical Chemistry · 10.11.2026 · 98 (44) · 30132-30140

DOI: 10.1021/acs.analchem.5c07805 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Analytical Chemistry, online erschienen am 14.10.2026. Verfasst von Yunze Li und 5 weiteren Autor:innen. Beteiligte Einrichtung: State Key Laboratory of Optical Fiber and Cable Manufacture Technology Guangdong Key Laboratory of Integrated Optoelectronics Intellisense Department…. Die Arbeit nennt 32 Literaturangaben. Gefördert von: Shenzhen Key Laboratory of Deep Subwavelength Scale Photonics, High-Level Special Funds, Shenzhen Science and Technology Program.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Raman hyperspectral imaging provides label-free chemical contrast, but quantitative analysis is limited by weak signals, fluorescence background, instrumental noise, and overlapping component spectra. Here, we present a simulation-driven cascaded attention U-Net (AU-Net) for Raman spectral denoising and targeted component unmixing directly from raw spectra. The training data were generated from measured pure-component references combined with randomized fluorescence baselines, nuisance Raman features, reference-spectrum perturbations, and mixed Poisson-Gaussian noise. This design provides paired raw inputs, denoised spectra, and component-level targets without requiring experimentally annotated mixture spectra. AU-Net was evaluated on simulated lipid-protein-DNA mixtures, sugar-mixture spectra, and HeLa-cell Raman hyperspectral images and was compared with NNLS, MCR-ALS, and an autoencoder-based Raman unmixing baseline under explicit preprocessing and reference-use assumptions. The benchmarks show that the raw-input neural workflow improves component recovery under the tested noise, baseline, and interference conditions, while additional stress tests define limitations under reference-spectrum variation, bilinear mixing, and changed target libraries. These results establish AU-Net as a reference-anchored workflow for repeated Raman analyses with a stable component library rather than a blind or universally transferable unmixing method.

Autor:innen und Institutionen

  • State Key Laboratory of Optical Fiber and Cable Manufacture Technology Guangdong Key Laboratory of Integrated Optoelectronics Intellisense Department of Electronics and Electrical Engineering, Southern University of Sci…
  • Institute of Nanoscience and Applications, Southern University of Science and Technology, Shenzhen518055, China.

ORCID-Profile: Yunze Li, Gina Jinna Chen

Förderung

  • Shenzhen Key Laboratory of Deep Subwavelength Scale Photonics
  • High-Level Special Funds
  • Shenzhen Science and Technology Program
  • National Key R&D Program of China
  • National Natural Science Foundation of China
  • Guangdong Basic and Applied Basic Research Foundation

Zitieren

APA

Li, Y., Hu, J., Lai, T., Wang, L., Shum, P. P., & Chen, G. J. (2026). A Simulation-Driven Deep Learning Approach for Artifact-Resistant Raman Spectral Unmixing. Analytical Chemistry, 98(44), 30132-30140. https://doi.org/10.1021/acs.analchem.5c07805

Vancouver

Li Y, Hu J, Lai T, Wang L, Shum PP, Chen GJ. A Simulation-Driven Deep Learning Approach for Artifact-Resistant Raman Spectral Unmixing. Analytical Chemistry. 2026;98(44):30132-30140. doi:10.1021/acs.analchem.5c07805

BibTeX
@article{Li2026,
  author = {Li, Yunze and Hu, Jiaqi and Lai, Tianyu and Wang, Lilian and Shum, Perry Ping and Chen, Gina Jinna},
  title = {A Simulation-Driven Deep Learning Approach for Artifact-Resistant Raman Spectral Unmixing},
  journal = {Analytical Chemistry},
  year = {2026},
  volume = {98},
  number = {44},
  pages = {30132--30140},
  doi = {10.1021/acs.analchem.5c07805}
}
Kostenlos starten →