IZSVe’s new technology to find chemical fingerprints with machine-learning software

Istituto Zooprofilattico Sperimentale delle Venezie (IZSVe, the Italian health authority and research organization for animal health and food safety) found a new way to detect oregano frauds with accuracy and precision. This is what the institution stated in a recent article published on Food Control, which focused on the use of a mass spectroscopy technology associated with machine-learning software.

 Oregano is one of the most commonly defrauded aromatic herbs, both accidentally and voluntarily. It is fundamental to develop new, more effective ways to detect food frauds in order to protect the consumer’s health and maintain the industry’s credibility. This new study used a direct analysis in real-time, high-resolution mass spectrometry (DART-HRMS) to gain a complex mass spectrum time-related. This spectrum, which is a sort of oregano’s chemical fingerprint, was analyzed by a statistical method based on a machine-learning model. In the first place, the software was “taught” to identify the spectrum of real oregano using previous standardized tests; after that, the machine was able to distinguish real oregano from adulterated one with 94% of accuracy.  

The authors declared that this evaluation technology is very promising because it is a non-targeted system, fast and accurate, and it could be used not only to find different food ingredients, but also food toxins. EFSA is incentivizing these technologies, and the Association of Official Analytical Collaboration (AOAC) recently stated in a document that non-targeted methods capable of detecting the chemical fingerprint of authentic food will be critical in detecting food frauds quickly.    

Source:

https://www.sciencedirect.com/science/article/abs/pii/S0956713521001961