Quantification of Saleable Meat Yield Using Objective Measurements Captured by Video Image Analysis Technology
ABSTRACT Video image analysis (VIA) images from grain-finished beef carcasses [n = 211; of which 63 did not receive zilpaterol hydrochloride (ZIL) and 148 received ZIL before harvest] were analyzed for indicators of muscle and fat to illustrate the ability to improve methodology to predict saleable meat yield of cattle fed and not fed ZIL. Carcasses were processed in large commercial beef processing facilities and were fabricated into standard subprimals, fat, and bone. Images taken by VIA technology were evaluated using computer image analysis software to quantify fat and lean parameters which were subsequently used in multiple-linear regression models to predict percentage of saleable meat yield for each carcass. Prediction models included variables currently quantified by VIA technology such as LM area (LMA), subcutaneous (SC) fat thickness at 75% the length of the LM (SFT75), and intramuscular fat score (IMF). Additional distance and area measures included LM width (LW), LM depth (LD), iliocostalis muscle area (IA), SC fat thickness at 25, 50, and 100% the length of the LM (SFT25, SFT50, SFT100), SC fat area from 25 to 100% the length of the LM (SCFA), and SC fat area adjacent to the 75% length of the LM from the spinous processes (SCFA75). Multiple ratio and product variables were also created from distance and area measures. For carcasses in this investigation, a 6 variable equation (Adj. R(2) = 0.62, MSE = 0.022) was calculated which included coefficients for ZIL treatment, SCFA75, LW, SCFA, SCFA/HCW, and SFT100/HCW. Use of parameters in the U.S. (Adj. R(2) = 0.39, MSE = 0.028) and Canadian [Adj. R(2) = 0.10, root mean square error (MSE) = 0.034] yield grade equations lack the predictability of the newly adapted equations developed for ZIL-fed and non-ZIL-fed cattle. Prediction equations developed in this study indicate that the use of VIA technology to quantify measurements taken at the 12th/13th rib separation could be used to predict saleable meat yield more accurately than those currently in use by U.S. and Canadian grading systems. Improvement in saleable meat yield prediction has the potential to decrease boxed beef variability via more homogeneous classification of carcass fabrication yield.
- [Show abstract] [Hide abstract]
ABSTRACT: Beef value is in the eye, mouth or mind of the consumer; however, currently, producers are paid on the basis of carcass grade. In general, affluent consumers are becoming more discerning and are willing to pay for both credence and measureable quality differences. The Canadian grading system for youthful carcasses identifies both lean yield and quality attributes, whereas mature carcasses are broadly categorized. Opportunities exist to improve the prediction of lean meat yield and better identify meat quality characteristics in youthful beef, and to obtain additional value from mature carcasses through muscle profiling. Individual carcass identification along with development of database systems like the Beef InfoXchange System (BIXS) will allow a paradigm shift for the industry as traits of economic value can be easily identified to improve marketing value chains. In the near future, developing technologies (e.g., grade cameras, dual energy X-ray absorptiometry, and spectroscopic methods such as near infrared spectroscopy, Raman spectroscopy and hyperspectral imaging) will be successfully implemented on-line to identify a multitude of carcass and quality traits of growing importance to segments of the consuming population.Canadian Journal of Animal Science 12/2014; 94(4):545–556. DOI:10.4141/CJAS-2014-038 · 0.98 Impact Factor