Application of NIR spectroscopy coupled with DD-SIMCA class modelling for the authentication of pork meat

Rearing system, breed, geographical origin, and animal welfare are some of the more recent quality features able to drive consumer's choice in meat and meat products purchase. The same aspects are marketing leverage from the producer's side. However, there is a need to develop rapid analytical strategies to check the compliance of declared quality with actual quality features. To face with this issue, this work aims at test the feasibility of using the fat portion as a marker of authenticity. To this end, fat samples from extensively and intensively reared pigs have been collected and analysed by NIRS (Near Infra-Red Spectroscopy). Then, after multivariate exploration, the Data Driven variant of Soft Independent Modelling of Class Analogy (DD-SIMCA) was used to develop models for the target class. The type of signal pre-processing and the number of principal components (PCs) used strongly affected the performance of the classification. Excellent results were obtained with SNV pre-treated data and 4 PCs that allowed to reach 100 % sensitivity and specificity in calibration and validation.