Evaluation of biomarkers that influence the freshness of beef during storage using VIS/NIR hyperspectral imaging

Abstract

Biomarkers influencing the freshness of beef during storage were detected using VIS/NIR hyperspectral imaging (HSI). A total of 18 cuts of eye round from three cattle were vacuum-packaged and wet-aged at 4 ± 2°C for 27 days. Throughout this period, freshness was maintained as evidenced by a significant decrease in pH, stable color, and total bacterial count (TBC) and volatile basic nitrogen (VBN) remaining below spoilage thresholds at 5.78 Log CFU/g and 14.47 mg/100 g, respectively. Metabolite profiling revealed correlations between freshness indicators-ethanol, 5’-inosine monophosphate, acetate, histamine-and TBC and VBN values, highlighting their importance in freshness. Integrating HSI with partial least squares regression (PLSR) proved more reliable than artificial neural networks for predicting metabolite profiles and correlating them with quality traits, confirming its effectiveness in meat quality monitoring. With PLSR, the model performance for TBC was similar (R 2 = 0.77 from HSI and 0.74 from metabolite predictions), while VBN performance improved significantly from R 2 = 0.63 to 0.81 with predicted metabolite data. This integration was essential for monitoring beef quality during wet aging and for developing effective assessment strategies. • VIS/NIR hyperspectral imaging was used to detect freshness biomarkers • Maintained beef freshness for 27 days detected by pH, color, TBC, and VBN • PLSR outperformed ANN in metabolites with R 2 : TBC = 0.77, and VBN = 0.81 • Four key metabolites correlated with TBC and VBN as freshness biomarkers • It was efficient metabolite predictions and quality traits for freshness control

Topics & Concepts