Peters delivered a presentation on a method she developed to identify poultry diseases far earlier than has been possible in the past. Altering an algorithm that was originally developed to detect deviations in industrial manufacturing, Peters created a method to identify very mild deviations in poultry production mortality data as an earliest possible warning that a disease might be developing in a poultry flock.

“The algorithm was originally created to detect problems in thread-spinning in industrial manufacturing,” says Peters. “We wanted a similar early warning tool, but that algorithm used uniform constants – thread spinning does not change throughout the production cycle. Because poultry mortality rates change through the life cycle of the animal, we decided to create a constant for every day of the life of the chicken. Once we had those constants, we could detect very mild changes from the average, which might point toward the first stages of a very big problem. And the great thing is that we can use data that poultry producers already collect as a matter of course.”

Peters conducted the project as a requirement for her thesis for an advanced degree in veterinary public health management. Dr. John Berezowski, veterinary epidemiologist with the food safety division supervised the project. Five farmers contributed data that they had collected over three years from about 250 flocks.

Dr. Rosemary Hood, of the Canadian Animal Health Surveillance Network (CAHSN), said she was impressed by the presentation. “Early detection is the key to a quick response to serious infectious diseases, she says. “Dr. Peters has implemented a data monitoring method that brings critical assessment and oversight to the day-to-day operations using a systems approach.” This approach could lead to more immediate investigations before the disease spreads.

Peters’ breakthrough research will play an integral role in the Alberta Veterinary Surveillance Network (AVSN).