A High-Resolution Workflow for Volatilome Profiling in Artisanal Cheese Using GC×GC-MS/FID and Image-Based Pattern Recognition

The quality of artisanal cheese is the result of complex interactions between raw materials composition, microbial consortia, ripening conditions and time. These factors shape a dynamic biochemical landscape that contributes to the sensory identity of the final product. However, conventional assessment methods, based primarily on sensory analysis and visual inspection, struggle to capture the molecular basis of cheese quality, limiting both their objectivity and diagnostic power. Modern approaches to food quality assessment increasingly rely on high-resolution analytical platforms capable of extracting and interpreting the chemical signatures embedded in complex matrices. The volatilome - the totality of volatile and semi-volatile metabolites - reflects both the metabolic activity of the fermenting microbes and the physico-chemical transformations during ripening and thus provide information about authenticity, process variability and quality attributes. In this study, a fully integrated analytical workflow was applied to investigate the volatilome of Valcasotto cheese, a traditional Italian dairy product (Prodotto Agroalimentare Tradizionale) [1]. The platform combines multiple headspace solid-phase microextraction (MHS-SPME) with comprehensive two-dimensional gas chromatography coupled with mass spectrometry and flame ionization detection (GC×GC-MS/FID). This configuration ensures high chromatographic resolution, broad chemical coverage and accurate quantification capability through a FID relative response factor (RRF)-based concept strategy [2,3]. To support the identification of compositional trends throughout the production chain, computer vision tools were used as an image pattern recognition layer. By translating chromatographic data into interpretable image profiles, this approach enabled rapid comparison of sample groups, from raw milk and curd to cheeses matured for 30, 90 and 120 days either under refrigerated conditions or traditional caves, based on their chemical fingerprints [4]. Volatile markers with high discriminant power were identified at different ripening stages, including acetoin, phenylethyl alcohol, γ-terpinene, and 1,8-cineole in curd samples, which are indicative of cow farming and milk quality and early fermentation dynamics. In intermediate to late maturation stages, particularly under traditional cave conditions, compounds such as 1-octanol, 1-octen-3-ol, δ- decalactone, ethyl hexanoate, dimethyl sulfone, and short-chain fatty acids (butanoic, 2- methylbutanoic, propionic acid) emerged as key differentiators. These molecules reflect progressive lipid catabolism, microbial succession, and environment-specific metabolic pathways. This work highlights the potential of high-resolution volatilomic workflows, integrated with image data analytics, to decode the molecular signatures underlying fermentation-driven transformations in artisanal dairy systems. The proposed platform provides a reproducible and scalable strategy for quality characterization and product valorization across traditional food supply chains.