Mid-infrared (MIR) spectroscopy is a rapid and non-destructive monitoring technique. Itmay offer unique information regarding cheese structure and hence ripening stage, which is critical toproduct quality. The objective of this study was to investigate the potential of MIR spectroscopy inconjunction with chemometrics to discriminate between Cheddar cheese samples based on ripeningage. Cheddar cheese samples (n = 24) were manufactured using either natural or bacterial enzymesand aged for 12 months at 8 °C. MIR spectra (640 – 4000 cm-1) were collected at 4, 6, 9 and 12months, pre-treated and subjected to principal component analysis. Using this method cleardifferentiation was observed in a principal component (PC) scatter plot (PC1 v PC2) between the 4/6,9 and 12 month samples. PC loadings plots indicated that the main regions of differencecorresponded with fat bands. The pre-treated data was also subjected to partial least squaresregression. Three models were developed. The first contained 4 ripening ages (4, 6, 9 and 12 month)and predicted age with an RMSEP of 1 month. The 4 and 6 month samples were combinedproducing a < 6 month sample group and the RMSEP fell to 0.73 months. The final modelincorporated just 6, 9 and 12 month samples and had an RMSEP of 0.66 months. Therefore this
展开▼
