Authentication of ripened sausages according to the animals rearing system by NIR hyperspectral imaging

Rearing system, breed, geographical origin are some of the 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, the compliance of a specific product with the declared - or supposed - quality features is mostly of a documentary nature, resulting in a possible distrust. Indeed, according to the Alert and Cooperation Network, meat and meat products are the third most reported category concerning non-compliance and suspicions fraud notifications, mostly related to faulty labelling or claims (European Union, 2023). Bearing this framework in mind, both authorities and consumers could profit from the development of rapid, non-destructive solutions to certify meat and meat products with claims of quality. In a previous study (Totaro 2023) it was shown that Near InfraRed Spectroscopy (NIRS), coupled with chemometrics, allowed to classify the rearing system of animals by using only the fat portion in unprocessed meats. However, when moving to processed products, minced and fermented such as ripened sausages, the things start to get complicated. Indeed, these products have usually highly non-uniform surfaces, with both fat and lean portions irregularly mixed and the analysis of the fat portion by single spot NIRS is often technically impossible. In the attempt to overcome this issue, in this work NIR HyperSpectral Imaging (HSI) was tested. A total of n = 12 different sausages obtained from both extensive (n = 7) and intensive (n = 5) declared rearing systems of the pigs were collected. The hyperspectral images were acquired using a HERA hyperspectral camera (Nireos SRL, Milan, Italy), in the range 900 - 1700 nm. A total of eighteen images were acquired (3 random positions × 3 salami slices × 2 replicates). The camera was controlled by the HERA ANALYSIS App software that, after proper corrections, allows to export the normalized images. These were imported and processed using Matlab R2021a (The MathWorks, Inc., Natick, MA, USA) and HYPER-Tools (Mobaraki 2018) version 3 (freely available at https://www.hypertools.org/; last accessed January 2024). After background removal and data exploration, K-means clustering was used to separate the pixels of each sample into two clusters (fat/lean) (Figure 1A). Then, the NIR signal of the sole fat pixels of each salami was averaged and preliminarily subjected to Principal Component Analysis (PCA) after proper preprocessing. By means of PCA a distinct clustering of the samples according to the declared rearing system was highlighted into the scores plot (Figure 1B), highlighting a great potentiality of the applied technique. As next step, an authentication model by Soft Independent Modelling of Class Analogy (SIMCA) will be calibrated and tested.