Predicting internal adipose from selected computed tomography images in sheep

Dairy cattle preferentially lay down internal (INT) (e.g., kidney) adipose relative to sub-cutaneous adipose (SCF) tissue, but this pattern is the opposite in beef cattle, which preferentially deposit adipose subcutaneously. It is proposed that these differential patterns of fat distribution evolved to match the different production drivers of these breeds. As the New Zealand sheep industry has become more diverse in terms of breeds and production systems, the role that differential adipose distribution could play in supporting different production systems is of interest. The potential to use commercially obtained Computed Tomography (CT) images (i.e., six images, selected to predict carcass attributes in the live animal), to additionally predict INTF in addition to carcass fatness, was investigated using a full set (i.e., 30-32) of CT images available on 50 maternal ewe lambs. The R2 between the fulland sub-set of images for absolute amounts ranged from 0.25 for bone to 0.92 for SCF, with INTF 0.77 (P<0.001). When adjusted for total weight, correlations were slightly reduced, but still supported the ability of the analysis of the six images to accurately identify animals with differential fat distribution. Future work will include analysis of a further 550 animals from the same data set to determine the genetic correlation between the full and subset of images. If successful, adipose distribution can be predicted from existing and future commercial CT data sets and used to investigate its genetic and phenotypic relationship with production traits.