Discrimination of normal and wooden breast chicken fillets using NIR, fluorescence and Raman spectroscopy

Abstract

Wooden breast (WB) is a concerning myopathy affecting broilers that results in hardened and pale fillets, with lowered pysicochemical, technological and textural traits. Near infrared spectroscopy (NIRS) has been successful to sort defective meat, but other spectroscopic methods, such as fluorescence emission, based on the detection of fluorophores, and Raman, based on inelastic scattering, have never been tested for this purpose. Breasts (40 normal, 40 WB) from a commercial slaughterhouse were selected by an experienced veterinarian and measured with NIRS (780–1080 nm), fluorescence emission (350–580 nm) after excitation at 330 nm, and Raman (100–3250 cm−1, 50 s of exposure moving the sample). Low-field nuclear magnetic resonance (LF-NMR) was used to measure T2 relaxation distribution, and water, fat and collagen content were measured for reference. Finally, PLS models assessed the discriminant power of each technology. The NIRS and NMR spectra showed more loosely bound water in WB. Fluorescence allowed to detect collagen and collagen crosslinking, as well as adipose tissue, and revealed two groups within the WB, one with greater collagen, and another with higher fat content. Raman spectra displayed a greater distinction of these two groups, and revealed less protein in WB, besides increased collagen and fat. Reference measurements of moisture, fat and collagen were well correlated with principal componentsof spectral data from each method, confirming the interpretation. Although NIRS accurately discriminates between normal and WB (100% accuracy), fluorescence, with 95%, and Raman, with 100%, revealed markers that could be used to assess the degree of both fibrosis and lipidosis in WB, providing a more detailed characterization of histological lesions.