Evaluation of visible spectroscopy and low-field nuclear magnetic resonance techniques for screening the presence of defects in broiler breast fillets

Abstract

The potential of using simple visible spectroscopy (VIS) and low-field nuclear magnetic resonance (LF-NMR) was evaluated for differentiating three muscle conditions, normal (NORM), pale, soft and exudative (PSE), and woody breast (WB), in broiler pectoralis major. A total of 162 of pectoralis major, 54 for each condition, were collected from a commercial plant. Meat VIS spectra and transverse relaxation curve were collected using a spectrophotometer and LF-NMR, respectively. Principal component analyses (PCA) showed the better separation of PSE from NORM by VIS spectra. LF-NMR distinguished WB from NORM and PSE meat. However, the combination of VIS spectra and NMR relaxation curve differentiated the three conditions. PCA linear discriminant analysis and support vector machine discriminant analysis models further demonstrated that PCA analyses of combined VIS and LF-NMR data can correctly identify the presence of the defects in broiler breast meat.