A new technique for baseline calibration of soil X-ray fluorescence spectra based on enhanced generative adversarial networks combined with transfer learning

文献情報

出版日 2023-09-28

DOI 10.1039/D3JA00235G

インパクトファクター 4.023

著者

Xinghua He, Yanchun Zhao, Fusheng Li

要旨

Obtaining accurate characteristic spectra and the net peak area is crucial in X-ray fluorescence (XRF) quantitative analysis. To improve the calculation accuracy of the net peak area of the characteristic X-ray spectrum, carrying out the baseline calibration is necessary before resolving the spectrum. First, this article proposes the use of an enhanced generative adversarial network (EGAN) depth network model for baseline calibration of XRF spectra. This method directly takes a clean spectrum after deducting the background as the optimal target for EGAN deep network training, thus eliminating the traditional cumbersome baseline fitting process. It can directly and quickly obtain the XRF spectrum after baseline calibration, and the entire process requires neither fitting the background nor knowing the background data. Further, to improve the generalization ability of the EGAN model, we introduce transfer learning in the model training process. We use the existing alloy sample spectrogram data to pretrain the EGAN model generator and then migrate the pretrained XRF-EGAN generator model to the baseline calibration task of 57 national standard soil sample spectra collected. During the model training process, a cross-validation method is used to train and test the model’s effectiveness. Finally, experiments are conducted on simulated spectra and actual XRF spectra to verify the accuracy and adaptability of the proposed method. Instrument calibration curves for peak counting and element concentration are established to verify the effectiveness of baseline calibration. The coefficient of determination R2 of the calibration curves for elements Cu, Zn, Mn and Cr is increased to 0.998, 0.988, 0.922 and 0.991, respectively. The results indicate that our proposed method can effectively estimate the pure spectrum after deducting the background. In addition, this method can also be applied to the baseline correction of other similar spectral signals.

掲載誌

Journal of Analytical Atomic Spectrometry

CiteScore: 6.2

自己引用率: 25.8%

年間論文数: 254

The Journal of Analytical Atomic Spectrometry (JAAS) is the central journal for publishing innovative research on fundamentals, instrumentation, and methods in the determination, speciation and isotopic analysis of (trace) elements within all fields of application. This includes, but is not restricted to, the most recent progress, developments and achievements in all forms of atomic and elemental detection, isotope ratio determination, molecular analysis, plasma-based analysis and X-ray techniques. The journal welcomes full papers, communications, technical notes, critical and tutorial review articles, editorials, and comments, in addition to the Atomic Spectrometry Updates (ASU) literature reviews that are prepared by an expert panel. Submissions are welcome in the following areas, but note this list reflects the current scope and authors are strongly encouraged to contact the Editorial team if they believe that their work offers potentially new and emerging research relevant to the journal remit: Fundamental studies in the following. New and existing sources for atomic emission, absorption, fluorescence and mass spectrometry and those that provide both atomic and molecular information Sample introduction techniques for solids, liquids, gases Improvements in sensitivity, selectivity, precision, accuracy and/or robustness Isotope ratio measurements, including techniques for improving precision and mass bias correction Single channel and multichannel simultaneous detection systems Chemometrics, statistics, calibration techniques and internal standardisation Theoretical and numerical modelling of fundamental processes related to all of the above methodologies Novel or improved methodologies in areas of application including, but not limited to the following. Biosciences, including elemental, speciation and isotopic analysis in biological systems, immunoassays based on metal-labeled antibodies, bio-imaging, and nanoparticle toxicology Geochemistry Environmental science Materials science, including engineered nanoparticles and quantum dots Metrology, including reference materials Forensic analysis Food and agricultural sciences Energy Archaeometry Molecular analysis. Molecular sources for elemental and isotopic analysis Atomic sources for molecular analysis Atomic and molecular techniques simultaneously used for complementary chemical information All contributions are judged on originality and quality of scientific content, and appropriateness of length to content of new science.