Building a new atlas for pig breeding


Building a new atlas for pig breeding

With the advancement of large-scale computed tomography (CT) in the Topigs Norsvin breeding system, there has been a huge increase in genetic progress in body composition and lameness detection.

The focus has been on lean meat percentage and yield (carcass to live animal), which treats the animal as a pool of digital pixels from CT images.

Moving ahead, atlas-based segmentation is a framework for performing dissection or cutting in a database of  live pig images. The construction of a pig atlas, or map for breeding, can be seen as the phenotypic analogue to mapping the genome in genomics.

Obtaining information from such non-invasive imaging techniques as CT requires relationships between the images and the physical reference to be studied. This relationship can be represented as an atlas by labelling different objects in the CT images based on density (fat, lean meat, bone), anatomy (liver, lungs, muscles) or different cuts of meat (belly, loin, ham, shoulder, etc.). The atlas will allow us to make further advances in the use of the database of CT images and its application in phenotypes that can be used in the Topigs Norsvin breeding system.

Read the article: