An international collaboration of researchers has developed an artificial intelligence (AI) program that has the potential to revolutionise lab research.
The team that built the program—Rutgers Office for Research leaders in New Jersey, working with others around the world—presented its findings in a paper published in Lab Animal.
Current methods of checking the state of animals, or assessing animal condition, have limitations. Lab technicians typically conduct assessments during daytime hours, when the animals are less active, and signs of health issues can be subtle and easy to miss.
The innovation the team developed improves the wellbeing of mice through 24/7 non-invasive tracking of a digital biomarker—locomotor activity—which is then analysed using a large language model (LLM) to provide alerts to veterinary staff about potential health issues. This AI algorithm detected potential health issues 3–5 days before obvious clinical signs appeared in mice, marking a milestone in health care for mice.
“The ability to find animals that are likely to get sick 2–3 days before visible symptoms present themselves provides us with an opportunity to really make a difference in improving the health of the animals,” Dr Jeetendra Eswaraka said.
“Improving veterinary care and improving operational efficiency by over 50 per cent are examples of how AI can change the world around us.”
