Quick method for assessing the quality of extra virgin olive oil

A team of researchers from the Universities of Bari and Milan has developed an innovative, non-destructive analysis system for the rapid, sustainable, and low-cost evaluation of extra virgin olive oil (EVOO) quality. This breakthrough represents an effective tool in the fight against adulteration and fraud in one of the most affected food sectors.

The new protocol, published in the journal Food Chemistry, is based on the application of chemometric and artificial intelligence (AI) approaches to Fourier-transform infrared (FT-IR) spectroscopy.

The main objective is to estimate the concentration of fatty acid ethyl esters, which are key indicators of EVOO quality and authenticity. Currently, this parameter is determined by gas chromatography (GC). Although GC is reliable, it is a complex, slow, and expensive procedure that requires the use of chemical reagents and specialized laboratories.

The proposed method takes advantage of the information obtained by FT-IR spectroscopy, which generates a spectral fingerprint of the product. This fingerprint is analyzed using multivariate analysis and machine learning models.

The most effective algorithm identified in the study uses the XGBoost technique. Also thanks to the use of explainable AI tools, the algorithm is capable of identifying and interpreting the spectral regions most strongly associated with the presence of ethyl esters. This process allows for the detection of correlations that are invisible to the human eye and traditional analysis methods, without the need to destroy the sample.

The implementation of this technology has the potential to drastically reduce analysis time and costs, as well as decrease the environmental impact associated with traditional methods. It will allow for rapid screening of larger quantities of samples, offering an immediate and reliable indication of product compliance.

Although this approach does not yet replace the official method (gas chromatography), it can immediately serve as an effective preliminary tool for producers, factories, consortia, and certification bodies, concretely improving quality control processes.

The research team, which is part of the METROFOOD-IT project, is already working to expand the experimental dataset and extend the methodology to the evaluation of other important EVOO quality parameters, such as:

 
• Acidity

 

• Peroxide value

 

• Phenolic content

The final goal is to develop an integrated system capable of offering a complete and rapid product assessment, revolutionizing quality control in the agri-food sector. The combination of spectroscopy with artificial intelligence is emerging as the future for making quality testing more accessible and efficient.