Correcting Multivariate Calibration Model for near Infrared Spectral Analysis without Using Standard Samples

Abstract

Model transfer is essential for practical applications of near infrared (NIR) spectroscopy because differences may exist between the spectra measured using different instruments. For correcting the calibration models in cases where standard samples are unavailable, a method is proposed based on the relationship between the prediction error and the spectral difference of the instruments. In this method, a partial least squares (PLS) model, named the primary model, was constructed using the calibration spectra measured with a primary instrument, and the model was then used to predict the spectra from a secondary instrument. Error may arise in the prediction and the predicted error is obviously caused by the difference between the spectra of the two instruments. Assuming that there is a linear relationship between the spectral difference and the prediction error, a correction PLS model can be built to predict the error using the spectral difference. The predicted error obtained with the correction model can be used to correct the error of the primary model. Two NIR datasets, of pharmaceutical tablets and tobacco leaf samples, measured with two instruments were used to test the performance of the method. The results show that, for both datasets, the prediction error for the spectra of the secondary instrument of the primary model can be satisfactorily corrected.

Modified linear model correction: A calibration transfer method without standard samples

“…Simultaneously, a comparison with the results obtained by the methods reported in our previous works 5,6 was performed. Table 1 lists the results (R 2 and RMSEP) obtained by correction model, 5 LMC 6 and mLMC (this work) methods, respectively with the different number of samples in the transfer set. In the calculations, the same transfer set and the same prediction set of the slave instrument spectra were used.…”

“…Therefore, the effect of the sample number of the transfer set was investigated. Simultaneously, a comparison with the results obtained by the methods reported in our previous works 5,6 was performed. Table 1 lists the results (R 2 and RMSEP) obtained by correction model, 5 LMC 6 and mLMC (this work) methods, respectively with the different number of samples in the transfer set.…”

“…Therefore, the dual model strategy was employed for model transfer when standard samples are unavailable. 5 The assumption in the work is that the difference in concentration of a component is proportional to the spectral difference for the samples with similar chemical and physical properties, even when the spectra were measured with different instruments. Therefore, a correction model can be built using the difference between the concentration of the predicting target and the spectra of different samples measured with two instruments, respectively.…”

Near Infrared Spectroscopy in China

“…Xiaoyong Li et al have come up with a new calibration transfer algorithm for NIR based on the relationship of the prediction error and the spectral difference between instruments. 3 The method assumes that there is a linear relationship between spectral differences between different spectrometers and the prediction error. Therefore, a correction model can be obtained with a group of reference samples.…”

Determination of Cadmium (II) in Aqueous Solutions by In Situ MID-FTIR-PLS Analysis Using a Polymer Inclusion Membrane-Based Sensor: First Considerations

“…In contrast, NIR-IR is characterized by the presence of broad bands and overtones, making this region less useful for univariate quantitative analysis; nevertheless, the significant differences in positions of functional groups also provide a source of information [ 27 ]. A useful way to solve the spectral overlapping and to use all the information contained in the spectrum of complex matrices is the multivariate calibration methods [ 28 ], extensively applied to NIR-IR [ 29 , 30 , 31 , 32 , 33 ]. Several works demonstrated that NIR-IR techniques can be applied as a tool for quantitative multivariate analysis, especially with the combination of separation and preconcentration methods for organic compounds [ 34 , 35 ], and metal ions [ 36 , 37 , 38 ], where some of them use NIR-IR followed by MID-IR and Raman [ 39 ].…”