An evaluation of the EagleTM FA DEXA (Dual-energy X-ray Absorptiometry) scanner as a method of estimating the chemical lean in cartons of boneless beef

The accuracy with which the EagleTM FA DEXA (dual-energy X-ray absorptiometry) scanner measured the chemical lean percentage (CL) of boneless beef in 27.2 kg cartons was evaluated under actual production operating conditions in a commercial meat plant. Forty cartons of boneless beef (CL range 60 to 98) were scanned in triplicate (Scan CL), and the results obtained were compared with chemically determined estimates of CL from three laboratories (Lab CL) performed in triplicate on each of three samples taken from the minced and blended contents of each carton. Differences between CL estimates from the different laboratories and from the different sub-samples were small and not statistically significant. The relationship between Scan CL and the Lab CL had an R2 value of 99.6% and a residual standard deviation (RSD) of 0.79. When estimates of errors associated with sampling and laboratory analysis (RSDs of 0.4 to 0.6) were taken into account, the Scan CL RSD was estimated to be between 0.6 and 0.7. The mean difference between Scan CL and Lab CL was 0.28. Thus the measures of accuracy were within the specifications for the scanner, which are stated as being an RSD of less than 1 and an average difference between Scan CL and actual CL of less than 1 percentage point. Estimates of CL obtained by analysing cores (about 1.5% of the total carton) taken from the cartons, were not closely related to Scan CL with an RSD of 2.00 for this relationship. Small but statistically significant differences were found between the CL groups (65, 80, 90, and 95%) in the deviations between Lab CL and Scan CL. It is concluded that the EagleTM FA on-line DEXA scanner provides a rapid on-line means of accurately estimating the CL content of boneless beef in standard cartons. (author). 11 refs., 2 figs., 4 tabs