V.TEC.1717 - DEXA and CT prediction of retail beef cut weights

Summary

By establishing the relationships between DEXA and CT measurements and beef cut weights from a range of retail cut specifications, this project will underpin the development of a Beef Value Calculator. This tool will allow the beef supply chain to predict the retail value of a carcases based on the composition of that carcase, the cut specifications, estimated processing costs, and the market values of the cuts specified. To ensure that the Beef Carcase Value Calculator is widely applicable to the beef industry, this bone out needs to be comprehensive – measuring all the saleable cut weights produced by the different cut specifications used in the Australian beef industry.

Objectives

Upon the completion of this project the relationship between DEXA and CT measurements and beef cut weights from a wide range of industry cut specifications will be determined by:
1. Development of a bone out protocol for beef
2. DEXA scan of beef sides calibrated to CT scan of beef primals
3. Generation of equations predicting cut weights from CT and DEXA LMY values and accuracy determined.

Key findings

In support of our hypothesis, the results demonstrate that in most cases DEXA variables can predict beef trim weights with higher precision and accuracy than models using the current measure of carcase fatness, P8 fat depth. This difference in prediction power was evident through small differences in precision indicators (R-square and RMSE), but also through the capacity of DEXA models to discriminate more variation in trim weight at a set carcase weight than models using P8 fat.

The results demonstrated that the commercial Beef DEXA system installed in Teys Lakes Creek abattoir has the capacity to improve the precision and accuracy of predicting beef cut weights. Though overall the precision (shown by R-squared and RMSE values) and accuracy (shown by bias and slope values) of cut weight predictions improved only a small amount with DEXA variables in the prediction models rather than P8 fat depth, the greater magnitude in cut weights differentiated by the DEXA models demonstrates the substantial difference in cut weights that can be realised with improved prediction models.

Benefits to industry

The results of this study support that the commercial beef DEXA system is able to provide an objective measure of carcase composition and thereby predict beef trim weights with better accuracy and precision than the current measure of carcase composition, P8 fatness. It also quantifies the additional mass of trim that can be discriminated by the DEXA system, enabling a clear business case to be established prior to installation of this technology.

MLA action

MLA continues to support the evaluation and accreditation of DEXA devices in both the beef and lamb industry with the aim of providing enhanced feedback for producers.

Future research

Future research will report on the ability of an on-line DEXA system to predict beef cut weights in a commercial setting, further quantifying the commercial viability of beef DEXA implementation into beef abattoirs within Australia and the potential returns on investment it will provide to processors and producers alike.