Prediction of technological and organoleptic properties of beef Longissimus thoracis from near-infrared reflectance and transmission spectra

Technological and organoleptic properties of beef cuts were predicted by near-infrared (NIR) spectroscopy. Spectra were collected on 189 beef Longissimus thoracis muscle samples using, transmission (NIRT) and reflectance with a probe. Quality assessment and NIR recordings were performed on sliced loin after 2 and 8 days ageing. Partial least squares regression yielded determination coefficients of cross-validation (R-cv(2)) of 0.12-0.25 for the prediction of Warner-Bratzler Peak Shear Force in reflectance and 0.15-0.41 in transmission. Higher R-cv(2) were obtained for L* parameter (0.83-0.85), a* (0.39-0.49) and b* (0.73-0.75) with reflectance. Predictions of drip loss and cooking loss were less accurate with a R-cv(2) range of 0.38 to 0.54 and 0.25 to 0.47, respectively. The NIR spectra collected on fresh meat show good potential to predict CIE L* and b* parameters in reflectance mode. (C) 2003 Elsevier Ltd. All rights reserved.