Detection of adulteration of cattle and buffalo meat through NIRS and chemometric analysis

Abstract

The objective of this study was to evaluate the efficacy of near-infrared (NIR) reflectance spectroscopy in detecting adulteration in cattle and buffalo meat. A total of 16 samples were tested, 2 of which were pure and 14 were adulterated. The beef samples were adulterated by mixing buffalo meat in the range of 0-28% (w/w) at approximately 2% increments according to weight. To detect adulteration, DLP® NIRscan™ Nano Software was used to gather spectra. The Unscrambler X program was used to develop calibration and validation models utilizing principal component regression and partial least squares. Root mean square error of calibration (RMSEc), root mean square error of cross-validation (RMSEcv), coefficient of calibration (R2 c ), and coefficient of cross-validation (R2 cv) were used to assess the accuracy of the calibration models. The R2 value of 0.90 or above indicates that the regression model is excellent. For the PCR model, the predicted R2 cv value was 0.73 and for the PLSR model, the predicted R2 cv value was 0.98 through leverage correction. In cross-validation, the R2 cv value was 0.65 for both the PCR and PLSR models. According to the findings, it is suggested that NIR spectroscopy is a reasonably efficient method for detecting adulteration in cattle meat with buffalo meat.