Comparison of transverse and longitudinal real-time ultrasound scans for prediction of lean cut yields and fat-free lean content in live pigs.

Live animal real-time ultrasound scans and carcass measures were taken on 80 pigs comprising two sexes (42 barrows; 38 gilts) and two halothane genotypes (40 carriers and 40 negatives) that were slaughtered between 108 and 148 kg live weight. Transverse scans (TRUS), at right angles to the midline, were taken on right (RS) and left (LS) sides at the last rib. Longitudinal scans (LON) were taken 6.5 cm off the midline immediately anterior (ANT) and posterior (PST) to the last rib on both the RS and LS. Longissimus muscle depth and area and backfat thickness over the longissimus muscle were measured on TRUS. Backfat thickness and longissimus muscle depth were measured at each end of the LON. Backfat thickness and longissimus muscle measurements were taken at the 10th and last rib on the RS of the carcass. Carcasses were fabricated using standard techniques to establish lean cut yields and carcass soft tissue was chemically analyzed to determine fat-free lean contents. Stepwise regression analysis was performed to develop equations to predict the weights and percentages of lean cuts and fat-free lean. Fat and muscle measures taken from ultrasound scans were generally less accurate than last rib carcass measures at predicting composition. There was little difference in R2 for equations based on either TRUS or ANT/LON; however, PST/LON, generally, were less accurate than ANT/LON. Combining measurements from more than one scan gave little improvement in R2 compared with the best single scan. Estimates of sex bias for carcass composition prediction were small. Halothane genotype and carcass lean content biases were detected; equations derived from the pooled data tended to overestimate the lean content of negative pigs and fatter carcasses and underestimate the lean content of carrier animals and leaner carcasses.