Using ultrasound technology to predict pork carcass composition.

Twenty market hogs were evaluated with real-time ultrasound both before and after slaughter. Fat measures (n = 9) were taken at various body locations along with the longissimus muscle area measurement at the 10th rib. After live ultrasound, the hogs were slaughtered and the unsplit carcasses were measured with ultrasound at the same live ultrasound locations. After chilling, carcass measures were taken at the same locations using a backfat probe for fat measures and a loin eye dot grid for measuring the longissimus muscle area. One side of each carcass was fabricated into the four lean cuts, which then were expressed as a percentage of the side weight. The most appropriate prediction equation found was a two-variable equation (fat thickness at the anterior tip of gluteus medius and longissimus muscle area) with a R2 of .83 and a RSD value of 1.67. This prediction equation was verified on a different sample of 20 market hogs; actual vs predicted four lean cuts revealed that the prediction equation had a R2 of .63 and a RSD value of 2.04. Although some accuracy and precision was lost when this live animal prediction equation was incorporated in market hog evaluation, this equation offers producers an objective mechanism for identifying carcass merit in live hogs.