Composition and intramuscular fat estimation of Holstein bull and steer rib sections by using one or more computed tomography cross-sectional images
2014
Font-i-Furnols, M. | Brun, A. | Marti, S. | Realini, C.E. | Perez, J. M. (Juan M.) | Gonzalez, J. | Devant, M.
Predictions of the composition and intramuscular fat (IMF) of beef rib sections were estimated from 1 cross-sectional image taken at the 10th rib level, from one at the 11th rib level (IMF) or from a series of cross-sectional images taken between the 9th and 11th ribs with computed tomography (CT) scanner. Two hundred and twelve beef cuts from Holstein bulls and steers were scanned, with two thirds of the samples used for calibration and the rest for validation. The 9–11th rib section was CT scanned and manually dissected. In addition, the IMF was determined in minced loin. For each scan, the volume associated with each Hounsfield value was obtained for the whole rib section and for the Longissimus thoracis muscle as regions of interest. Volumes associated with a range of Hounsfield values were selected in a stepwise manner and used as predictors in the regression equation. With the exception of IMF prediction, the accuracy was higher when the entire rib section was scanned in comparison to a single cross-sectional image taken at the 10th rib level. However, as one single image is cheaper than scanning the entire rib section, a compromise between precision and cost must be considered to select an adequate prediction method using one cross-sectional image or a series of images comprising the whole rib section. The CT results from IMF content determinations of L. thoracis muscle improve when one cross-sectional image of all the cut (including other muscles and types of fat) is considered, probably because the high correlation between IMF content and fat content of the whole section. Moreover further studies are needed to improve CT scanning predictive accuracy.
Показать больше [+] Меньше [-]Ключевые слова АГРОВОК
Библиографическая информация
Эту запись предоставил National Agricultural Library