期刊
ANIMAL
卷 9, 期 1, 页码 166-178出版社
CAMBRIDGE UNIV PRESS
DOI: 10.1017/S1751731114002237
关键词
computed tomography; live growing pig; carcass composition; cut composition; prediction equations
资金
- Instituto Nacional de Investigaciones Agrarias -INIA [RTA2010-00014-00-00]
The aim of the present work was (1) to study the relationship between cross-sectional computed tomography (CT) images obtained in live growing pigs of different genotypes and dissection measurements and (2) to estimate carcass composition and cut composition from CT measurements. Sixty gilts from three genotypes (Duroc x (Landrace x Large White), Pietrain x (Landrace x Large White), and Landrace x Large White) were CT scanned and slaughtered at 30 kg (n = 15), 70 kg (n = 15), 100 kg (n = 12) or 120 kg (n = 18). Carcasses were cut and the four main cuts were dissected. The distribution of density volumes on the Hounsfield scale (HU) were obtained from CT images and classified into fat (HU between -149 and -1), muscle (HU between 0 and 140) or bone (HU between 141 and 1400). Moreover, physical measurements were obtained on an image of the loin and an image of the ham. Four different regression approaches were studied to predict carcass and cut composition: linear regression, quadratic regression and allometric equations using volumes as predictors, and linear regression using volumes and physical measurements as predictors. Results show that measurements from whole animal taken in vivo with CT allow accurate estimation of carcass and cut composition. The prediction accuracy varied across genotypes, BW and variable to be predicted. In general, linear models, allometric models and linear models, which included also physical measurements at the loin and the ham, produced the lowest prediction errors.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据