Fast reconstruction of Raman spectra based on global weighted linear regression

https://doi.org/10.1016/j.chemolab.2020.104073Get rights and content

Highlights

  • We propose a method for optimizing the training samples based on Tanimoto coefficient to delete the bad sample.

  • We perform polynomial regression on the multi-channel measurements in order to reduce the influence of nonlinear factors and introduce the normalization, cross-validation algorithms to improve the accuracy of reconstructed spectra.

  • We establish a linear regression function by optimizing the training samples and making a global assignment weight to the optimizing samples.

  • We verify the effectiveness and the feasibility of the methods by processing the Raman spectra of several materials.

Abstract

Raman spectroscopy has shown great potential in biomedical applications. However, slow data acquisition of Raman spectra has seriously hindered the expansion of its application. In this paper, we have completed the reconstruction of Raman spectra with multi-channel measurements based global weighted linear regression. This algorithm establishes a linear regression function by optimizing the training samples and making a global assignment weighted to the optimizing samples. Simultaneously, the normalization and polynomial regression are introduced in order to improve the accuracy of reconstructed spectra. It has evaluated the Raman spectra of several materials. According to the root mean square error, the fitness between reconstructed and original spectra is excellent. This algorithm can be used in quickly testing for potential sample component in a substance, where the sample component to be tested is known and provides a theoretical support for the application of Raman imaging technology in fast dynamic systems.

Introduction

Raman spectroscopy has the advantages of no damage, no marking and no interference from water [1,2], etc. It can obtain the spectral information on tissues or cells under the close physiological conditions. Therefore, it has become a diagnostic or analytic tool for various diseases [3,4]. However, slow data acquisition of Raman spectra has seriously hindered the expansion of its application. In order to improve the speed of data acquisition of Raman spectra and the high resolution of Raman imaging, the researchers have done a lot research.

In order to settle this problem, several methods [5] have been explored. The first one is line scanning: a laser spot is concentrated on the surface of the sample by scanning along a line and the corresponding Raman spectral signals are imaged on the CCD of the spectrometer. It has achieved the purpose of simultaneously collecting multiple spectra, and effectively improved the speed of scanning. However, it is still time-consuming. Simultaneously, the resolution of Raman imaging can only reach the level of a few microns. The other one is Wide-Field imaging. By exciting the sample in a Wide-Field, the Raman scattering is directly coupled to the surface of CCD, and wavelength of the passed signal is converted to single wavelength imaging by using a Liquid crystal tunable filter (LCTF) [6]. Compared to line scanning, Wide-Field imaging can achieve a higher resolution and a better dynamic performance. However, it can only collect spectra of single wavelength at two spatial dimensions.

Therefore, it is important to develop an algorithm which is cost-effective and fast in data acquisition of Raman spectra. In this paper, an algorithm based on the fast reconstruction of Raman spectra has been put forward: Firstly, we get the Raman spectra set R of the sample A and the corresponding multi-channel measurements set U. Secondly, the transform Matrix W between R and U is established. Finally, the Raman spectrum of sample A is reconstructed by the multi-channel measurements u and W. All the above are shown in Fig. 1. In the process of fast reconstruction of Raman spectra, the solution to the transform matrix W is particularly critical. In 2005, D. R. Connah etc. [7] applied polynomial regression to multi-spectral imaging systems and achieved rapidly reconstruction of spectral reflectance. In 2013, Shuo Chen etc. [8] proposed a fast reconstruction of Raman spectra from narrow-band measurements based on Wiener estimate.

In the above algorithms, the solution to the transform matrix W, either ignores the influence of nonlinear factors, or disregard the optimization of the training samples. Therefore, a new algorithm is proposed in this paper. Firstly, the polynomial regression is performed on the multi-channel measurements in order to reduce the influence of nonlinear factors. Secondly, a linear regression function is established by optimizing the training samples and making a global assignment weighted to the optimizing samples. Finally, the reconstruction of Raman spectra is quickly achieved, by introducing the normalization and cross-validation algorithms. In addition, the quality of the reconstruction of Raman spectra will be determined based on the root mean square error (RMSE).

Section snippets

Reconstructed spectra based on global weighted linear regression

The Raman spectra set R of sample A (A is shown in Fig. 1) is obtained by taking a conventional Raman spectrometer(composed of QE65Pro, Raman fiber probe, and 785 nm laser illumination), which are used as training samples. The corresponding multi-channel measurements set U are achieved by calculating the inner product between each Raman spectrum of R and the spectral response function matrix of the band-pass filters at the wavenumbers of the characteristic peak. And so far, the preparation for

Spectral preprocessing

In this paper, ethanol and lactic acid are used as experimental samples. Measuring the concentration of 50% ethanol solution by a conventional Raman spectrometer and obtaining a training sample set of ethanolR={r1r2 ..., ri..., rk}, k represents the number of measurement. The concentration of 45%, 55% ethanol solution and lactic acid solution are measured ten times. Taking the best one from ten as the spectra to be detected and setting asr45,r55andrla, respectively. In order to improve the

Conclusion

In this paper, a fast algorithm for the reconstruction of Raman spectra based on Global weighted linear regression from multi-channel measurements is developed. This algorithm improves the accuracy of reconstruction by optimizing the training samples and making a global assignment weighted to the optimizing samples. According to the RMSE, it proves that this algorithm is 78.2% higher than the Pseudo-inverse Method in the accuracy of the Raman spectrum reconstructed. Simultaneously, it can be

CRediT authorship contribution statement

Xian-guang Fan: Resources, Supervision. Long Liu: Conceptualization, Methodology, Validation, Formal analysis, Writing - original draft, Writing - review & editing. Zhe-ming Kang: Investigation. Ying-jie Zeng: Writing - review & editing. Yu-liang Zhi: Writing - review & editing. Ying-jie Xu: Writing - review & editing. Jia-jie Zhang: Writing - review & editing. Xin Wang: Resources, Funding acquisition, Project administration, Supervision.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgement

The authors would like to acknowledge financial support from Natural Science Foundation of China (No.21874113).

References (13)

  • Y.H. Ong et al.

    Comparison of principal component analysis and biochemical component analysis in Raman spectroscopy for the discrimination of apoptosis and necrosis in K562 leukemia cells

    Optic Express

    (2012)
  • Y. Li et al.

    Micro-Raman spectroscopy study of cancerous and normal nasopharyngeal tissues

    J. Biomed. Optic.

    (2013)
  • C. Krafft et al.

    J. Biophot.

    (2009)
  • I. Nabiev et al.

    J. Raman Spectrosc.

    (2005)
  • Sebastian Schlucker et al.

    Raman microspectroscopy:A Comparison of point, line, and wide-field imaging methodologies

    Anal. Chem.

    (2003)
  • N. Everall

    Depth profiling with confocal Raman microscopy

    Part I. Spectroscopy

    (2004)
There are more references available in the full text version of this article.

Cited by (1)

View full text