Detection of minced lamb and beef fraud using NIR spectroscopy

摘要： In this study, the feasibility of NIR spectroscopy to detect different types of meat fraud in both minced lamb and beef was investigated. For this, a multivariate chemometric approach was used to identify the most useful pre-processing techniques to discriminate between pure lamb and beef and adulterated samples. The results obtained in this study suggest that it is possible to use NIRS to distinguish pure from adulterated minced meat with acceptable precision and accuracy. Rates of classification between 78.95 and 100% were achieved for the validation sets. Higher % CC samples were obtained for samples mixed with pork, meat of Lidia breed cattle and foal meat than for samples adulterated with chicken, where the lowest rates of classification were achieved in both lamb and beef. Additionally, identification of adulteration of meat of Lidia breed cattle in minced beef at 2% was achieved. Furthermore, best classification results were obtained for minced beef mixed with foal meat with a 100% of samples correctly classified indicating that inclusion of foal meat in minced beef at 1% and higher can be detected by using NIRS. Regarding pre-processing techniques, in general, the most powerful ones to classify both groups of samples (pure and mixed) were those orientated to reduce the scatter, MSC and SNV, and, those to correct peak overlaps, 1st and 2nd Der.