Comparison of four real-time ultrasound systems that predict intramuscular fat in beef cattle.

Eighty-one crossbred steers were used to evaluate four commercially available ultrasound systems that predict intramuscular fat. The software systems represented included Animal Ultrasound Services, Inc., Ithaca, NY; CPEC, Oakley, KS; Critical Vision, Inc., Atlanta, GA (CVIS); and Classic Ultrasound Equipment, Tequesta, FL. Systems were evaluated using marbling scores and percentage ether extractable fat of the longissimus muscle. Before statistical analyses, system predictions were corrected for the respective system's average deviation between the prediction and carcass measurement. The absolute difference between system prediction and percentage ether extract (EEADIFF) or marbling score (MADIFF) converted to ether extract by regression was analyzed with a model accounting for effects of system, technician within system, animal, and animal x system. Steers with USDA marbling scores less than Small00 were assigned to a low marbling class, and all others were assigned to a second class. Data were then analyzed with a subsequent model including marbling class. For EEADIFF in the first model, system, animal, and system x animal were significant (P < .001). For MADIFF, technician within system (P < .05) and all other effects (P < .001) were significant. In the second model, system x marbling class was significant (P < .05) for EEADIFF and approached significance for MADIFF (P = .17). Least squares means for EEADIFF indicated that the systems were more precise measuring animals in the low marbling class. Finally, the CPEC and CVIS systems were the most precise for predicting intramuscular fat.