Use of video image analysis, ribeye grids, and linear ribeye Use of video image analysis, ribeye grids, and linear ribeye measurements to predict and compare ribeye areas from carcass measurements to predict and compare ribeye areas from carcass left and right sides left and right sides

Ribeye tracings from 265 beef carcasses were used to compare ribeye areas from right and left sides. When video image analysis (VIA) was used to determine ribeye area, no difference (P=.48) was observed between right and left sides. However, when ribeye area was determine d by using USDA grids, those on the left side were slightly larger (P<.01) than those on right side. This difference is negligible considering the wide rang e in variation (SD=.


Introduction
Ribey e area is the muscling factor used in calculating yield grade.However, ribeye areas from the left and right sides of a carcass may differ.This study evaluated the difference in size of ribeyes between the left and right sides.Ribeye measurements are traditionally performed by using a USDA ribeye grid.However, at current chain speeds in most packing plants, the time needed to accurately "grid" a ribeye is too long.Many techniques have been used to accelerate collection of carcass data.Systems that measure images using current computer and video technology (video image analysis, VIA) offer faster and more accurate ribeye area measurements.In addition, simple linear measures of ribeye dimensions potentia l ly could predict ribeye area very rapidly if less accuracy is acceptable.Therefore, our objective wa s to compare USDA ribeye grids, linear measures, and VIA for determining ribeye area.

Experimental Procedures
Ribeyes f rom the right and left sides of 265 beef carcasses were traced onto acetate tracing paper at a commercial packing facility.Member s of the KSU Meats Judging Team measured the ribeye tracin gs using USDA grids.Two individuals measured each ribeye twice and averaged their two measurements.The measurement s from the two individuals were average d to determine the final ribeye area.If the difference between the two individuals' results was greater than 0.5 i n, a third person 2 measure d the ribeye and the measurement furthest from the mean was deleted.In addition, ribeye length, center mid-width, and widest width were measur ed using a ruler calibrated to the nearest .05in.Ribeye tracings also were measure d by VIA.Correlations, paired t-tests, and regression analysis were conducted on data.

Results and Discussion
A plot of right vs. left side ribeye areas using VIA is displayed in Figure 1.The ribeye areas ranged from 7.6 to 19.9 i n.When 3.4 in ) when the left side was larger.A paired t-test revealed no difference (P=.48) between right (12.92 in 2 ) and left (12.95 in 2 ) sides when measured by VIA.However, when USDA grids were used, ribeyes from the left sides (13.05 in 2 ) were slightly larger (P<.01) than ribeyes from the right sides (12.86 in 2 ).
Selected correlations of right and left side ribeye areas and measurements using USDA grids, a ruler, and VIA are presented in Table 1.Correlations between grid and VIA ribeye areas and between ruler and VIA-measured ribeye lengths were high for both the right (.96 and .97,respectively) and left (.95 and .97,respectively) carcass sides.For both right and left carcass sides, linear measures (length, midwidth, widest width, and VIA length) had moderately high correlations (.73 to .80) with ribeye area measures (Grid and VIA).However, length measures (Ruler and VIA) had lower correlations (.37 to .49) with width measures (mid-width and widest width).
For carcass right sides and pooled right and left sides (Table 2), ribeye areas measured with a grid or VIA were similar (P=.33).For the left sides, ribeyes measured with a USDA grid were slightly larger (P<.01) than VIA-measured ribeyes.However, the .1 in 2 difference is negligible when calculating yield grade.Measurement of ribeyes by either VIA or USDA grids can accurately determine ribeye area.Length of ribeyes measured by a ruler and VIA were similar (P=.51) for carcass left sides.For carcass right sides and combined right and left sides, ribeye lengths measured by VIA were slightly greater (P<.05) than those measured by ruler.Again the .05and .02in.differences are minimal and have little consequence compared to the wide range in ribeye lengths.
Regression equations (Table 3) were developed to predict VIA ribeye area from linear ruler measurements of ribeye length, midwidth, and widest width.Equations utilizing ribeye length and either mid-width or widest width had R 2 between .84 and .87 for right, left, and combined right and left side ribeyes.Equations combining ribeye length, mid-width, and widest width improved the R 2 to .90 and .91.Potentially, linear measurements could be collected at chain speeds of commercial plants by a data collection team.By incorporating these measures into regression equations, ribeye areas could be predicted with reasonable accuracy.This is especially true considering the differences that may exist between right and left side ribeyes.

Figure 1 .
Figure 1.Plot of Right Side vs. Left Side Carcass Ribeye Areas Using Video Image Analysis.