Potential of a custom-designed fluorescence imager combined with multivariate statistics for the study of chemical and mechanical characteristics of beef meat

The potential of fluorescence imaging to discriminate different bovine muscles in relation with animal age, muscle type, chemical and mechanical properties was examined. Twenty-four muscles of three types (Gluteus medius, Longissimus dorsi, and Semitendinosus) and two animal age groups (10–13-years old and 12–24-months old) were obtained from the carcasses of Limousin breed cows. One hundred and forty-four images were collected at three illuminating conditions (exc 320 nm, exc 380 nm, and white light) using a custom-designed imager. The image cubes were processed using “regionprops” algorithm developed earlier in order to extract image shape features (number of shapes, area, major-axis-length, eccentricity, and solidity). Extracted image shape features were processed using custom-designed programs. The results of the PLSDA performed on image shape features showed 100% good discrimination for the three types of muscles. Muscle samples were also subjected to chemical analysis (dry matter, fat, pyridinoline, total, insoluble and soluble collagen) and mechanical tests (shear stress and breaking energy). PLSR models indicated relations between extracted image shape features and mechanical properties, i.e., R2 = 0.69 and RMSEV = 0.514 were observed for breaking energy for adult-animal muscles. Regarding chemical composition, image shape features allowed to predict total collagen of L. dorsi with R2 = 0.61 and RMSEV = 0.756. This study has demonstrated a promising potential of the custom-designed fluorescence imager combined with multivariate statistical tools in the study of beef meat.

► Potential of fluorescence imaging to discriminate bovine muscles was examined.
► Image features were processed with PLSDA, CCSWA, and PLSR muscle properties.
► The study showed high potential of the fluorescence imager in the study of meat.