Artificial Intelligence decision-making tools based on comprehensive two-dimensional gas chromatography data: the challenge of quantitative volatilomics in food quality assessment

Abstract

Effective investigation of food volatilome by comprehensive two-dimensional gas chromatography with parallel detection by mass spectrometry and flame ionization detector (GC×GC-MS/FID) gives access to valuable information related to industrial quality. However, without accurate quantitative data, results transferability over time and across laboratories is prevented. The study applies quantitative volatilomics by multiple headspace solid phase microextraction (MHS-SPME) to a large selection of hazelnut samples (Corylus avellana L. n = 207) representing the top-quality selection of interest for the confectionery industry. By untargeted and targeted fingerprinting, performant classification models validate the role of chemical patterns strongly correlated to quality parameters (i.e., botanical/geographical origin, post-harvest practices, storage time and conditions). By quantification of marker analytes, Artificial Intelligence (AI) tools are derived: the augmented smelling based on sensomics with blueprint related to key-aroma compounds and spoilage odorant; decision-makers for rancidity level and storage quality; origin tracers. By reliable quantification AI can be applied with confidence and could be the driver for industrial strategies.