Detection of beef aging combined with the differentiation of tenderloin and sirloin using a handheld NIR scanner

There is an expressed need for non-destructive userfriendly tools that can help customers and various stakeholders of the food market to identify and qualify samples rapidly and accurately. The identification of high quality meat cuts and the determination of aging are important challenges where handheld near infrared spectroscopy can provide perfect solutions. The objective of this study was to develop multivariate models for differentiation of beef cuts and prediction of the aging time based on the NIR spectra acquired with a handheld Tellspec Enterprise Food Sensor. Sirloin and tenderloin samples were stored at 4°C in plastic bags for 10-day period during two experiments, and spectra were recorded daily. The investigated sirloin and tenderloin samples were separated in principal component analysis, and it was possible to use the principal components in a supervised classification (linear discriminant analysis) to build model on meat authentication. 85.37 % of the sirloin and tenderloin samples were classified correctly in independent validation tests. Multivariate calibration on aging was developed for the separate meat types. After omitting the first and last days of the experiments, accurate calibration models were built on the aging of beef samples. Accordingly, 1.1 or 1.5 days of precision was achieved during independent predictions for aging time of sirloin or tenderloin, respectively. Our results proved that the Tellspec Enterprise Food Sensor provides the possibility for rapid and non-destructive determination of meat type and stage of aging.