Using retinal images to track livestock
While livestock managers rely on traditional animal tracking methods, like ear tags and electric transponders, researchers in Spain have an eye on retinal imaging. A team of scientists from the Universitat Autònoma de Barcelona recently used retinal images (RI) to trace a group of lambs from weaning to yearling stages. With a 100 percent matching rate, they believe RI could be a successful addition to traditional identification methods.
In the experiment, researchers scanned 58 Manchega and Lacaure lambs. They took RI from the lambs using an OptiReader, a commercially available device that captures retinal vascular images in animals. The OptiReader can also store the location, time, and date of the imaging. “The Optibrand’s Data Management software overlaps the images to compute a MS [matching score] which ranges between 0 and 100. The greater the score, the more likely the images in the pair are from the same eye,” wrote the researchers in their paper, published February by the Journal of Animal Science.
Retinal vessels can change slightly as lambs age, so the team captured RI at three, six, and 12 months of age. The images taken at three months old were then used as the reference images. The researchers found that not only could OptiReader match two eyes from the same lamb, the software was accurate enough to trace a lamb using just one eye for identification. “Retinal imaging was an accurate technique for auditing the identity of living lambs,” they wrote.
Though the experiment was successful, the team did face challenges.
First, it took time to train the operator of the OptiReader device. Researchers had to restrain the lambs, then, like with any camera, OptiReader sometimes captured blurry images. Eventually, operators averaged about one minute per eye. “Operator training has been shown to be a key factor for collecting quality RI using the OptiReader device,” wrote the team. They recommended that future researchers take many practice images.
There was also a glitch with the OptiReader software. In 15 cases, the program reported a failure to match a pair of eyes. Upon investigation, the team found that these failures were false negatives, and the eyes did match. To prevent this mistake in the future, the researchers recommend manually checking rejected pair of images.
Overall, RI could be a good alternative to the limitations of traditional ear tags and transponders. The system is so accurate, even clones can be differentiated by RI. In addition, RI data can maintained on electronic databasesfor future reference.
The paper is titled “Retinal image recognition for verifying the identity of fattening and replacement lambs.” It can be read in full at http://journalofanimalscience.org/.
Scientific Communications Associate