Comparison of AOAC and light spectroscopy analyses of uncooked, ground beef

Ground beef samples (n = 302) were analyzed for DM and fat content with standard laboratory procedures. Samples were also analyzed by light spectroscopy with either a NIRSystems model 6250 or a model 6500. Samples were thawed at room temperature, packed in cups (three/sample), and scanned every 2 nm with the model 6250 in the mid wavelength (MW) region from 1,100 to 2,500 nm (MW-6250). Whirl-pak bags containing the thawed samples were also scanned every 2 nm with the model 6500 using a fiber optic attachment in the short wavelength (SW) region from 400 to 1,100 nm (SW-6500), 1,100 to 2,500 nm (MW-6500), and 400 to 2,500 nm (SW+MW-6500). Each sample was scanned at two different locations on the Whirl-pak bags. The MW-6250 produced the lowest standard errors of calibration (1.62 and 1.62%) and highest R2 values (.97 and .98) for DM and fat, respectively, during calibration. However, SW-6500 produced the lowest standard errors of performance (2.34 and 1.92%) and the highest r2 values (.86 and .95) for DM and fat, respectively, during validation. The MW-6500 and SW+MW-6500 produced intermediate results. Results with fiber optic scanning in the SW region were better than with cups and scanning in the MW region. Light spectroscopy for determining DM and fat content of ground beef samples can replace wet chemistry.