Application of the normalized relative error distribution analysis for non‐destructive quality control of drugs by Raman spectroscopy

Abstract

Development of drug quality control methods using non‐invasive methods, in particular Raman spectroscopy, is an important task in the pharmaceutical industry. This paper presents a new approach to the analysis of drug spectra obtained with portable Raman spectrometers. It represents a normalized relative error distribution analysis (nReDA). The proposed method uses the Pearson residuals distribution statistics and, compared with the correlation method, allows increasing the screening sensitivity to foreign components that are not contained in the original sample, with the same resistance to changes in the noise distribution and the peak ratios. On the basis of the statistical data, the nReDA in some way is similar to the techniques such as principal component analysis, however, it requires a considerably less spectra to build a representative model and is suitable for comparing Raman spectra, having a noise other than the Gaussian one. The paper shows the prospects of the algorithm for creating a spectral library and solving the problem of transferring Raman libraries to a large number of spectrometers of the same type.

Development of a low‐cost food color monitoring system

“…A comparison between the performance of the webcam and Picamera suggests that Picamera shows better results with a lower %relative error (less than 5%). Prior literature suggests that %relative error can be an efficient measure to evaluate the performance of a color monitoring device and has been used in various applications such as quality control of drugs, 32 quantitative chemical analysis, 33 foodstuff quality measurement, 34 pH mapping, 35 and so forth. Hence, the current study uses %relative error as the measure of performance of the device.…”

Recent advances in linear and nonlinear Raman spectroscopy. Part XIV

“…They used a confocal Raman microscopy to compensate for the nonhomogeneous distribution of keratin. Galeev et al [ 204 ] presented the application of the normalized relative error distribution analysis for nondestructive quality control of drugs by Raman spectroscopy. They show the prospects of their algorithm for creating a spectral library and solving the problem of transferring Raman libraries to a large number of spectrometers of the same type.…”

Adaptive noise model based iteratively reweighted penalized least squares for fluorescence background subtraction from Raman spectra

“…[ 6,7 ] Raman spectroscopy technology has been successfully implemented in portable devices that makes it easy to perform on‐site or field analysis by comparing the spectrum of the tested sample with the model one. [ 8 ] However, in some cases, the spectra may have a rather high fluorescent background caused by the characteristics of the sample or its packaging that often affects the work. To simplify the interpretation of such spectra, determination of peak position, and estimation of peak width and intensity, it is necessary to use the baseline removal and the fluorescent background subtraction algorithms.…”