Estimation of lean meat percentage in pig carcass with the use of objective methods with regard to sex

In the Czech Republic, the pig carcass classification is mandatory in slaughterhouses processing over 200 pigs weekly. As breeding practices evolve to enhance lean meat yield, it is essential to update regression equations used in classification systems. This study presents new regression models for the Fat-O-Meater II (FOM II) device, using computed tomography (CT) as the reference method. Separate equations were developed for barrows, gilts, and boars to improve the accuracy of lean meat percentage (LMC) estimation. To calibrate the CT method, 24 carcasses were selected across a range of backfat thicknesses and sexes. CT scans were performed on chilled left carcass halves, followed by manual dissection to determine the true LMC. A correction model was applied to align the CT-derived LMC with dissection results. Subsequently, 128 carcasses were measured using FOM II and CT to develop sex-specific regression equations using ordinary least squares. The models revealed sex-specific differences in prediction accuracy. Gilts achieved an R2 of 0.66 and RMSEP of 1.35; barrows had higher R2 (0.759) and greater RMSEP (1.46); boars showed the most consistent composition (R2 = 0.734, RMSEP = 1.14). Compared to the standard method, gilts and boars had slightly higher LMC (+0.03% and +0.82%), while barrows had lower LMC (-0.14%). These differences translated into economic impacts, with gains of CZK 1.23 and CZK 4.33 per gilt and boar carcass, respectively, and a loss of CZK 5.55 per barrow carcass. These results support the formulated hypotheses, and the fact that sex-specific calibration enhances classification accuracy and economic efficiency.