The current version of the Tecniplast website doesn't match your region. Please visit your local website to find information and offerings specific to your country.
The current version of the Tecniplast website doesn't match your region. Please visit your local website to find information and offerings specific to your country.
A new multicentric study published in Lab Animal demonstrates how artificial intelligence can enhance laboratory animal welfare by enabling earlier detection of health and welfare issues through continuous home-cage monitoring.
Conducted in collaboration with Rutgers University, Sanofi, EPFL, and Tecniplast, the study analyzed more than 156,000 cage evaluations collected across three institutions operating under different husbandry models. The research leveraged data generated by the Digital Ventilated Cage (DVC®) platform, which continuously records cage-level locomotor activity using non-invasive capacitance-based sensors installed beneath each cage.
Traditionally, animal welfare assessments rely on routine visual inspections performed by trained personnel. While essential, these checks typically last only a few seconds per cage and often occur during the animals’ inactive phase, making subtle behavioral changes difficult to detect at an early stage.
The study evaluated a machine learning algorithm capable of comparing each cage's activity patterns against its own seven-day baseline. By continuously analyzing behavioral data, the system identifies deviations such as hypoactivity, hyperactivity, unusual spatial activity patterns, or complete inactivity, generating alerts that can support veterinary and animal care staff in their decision-making process.
The results highlight the potential of
Rather than replacing human expertise, AI functions as an additional layer of continuous observation, monitoring every cage around the clock without disturbing the animals. This combination of human care and automated data analysis creates new opportunities to improve welfare assessment and support timely interventions.
The implications extend beyond animal care. Earlier detection of behavioral changes can help researchers identify potential confounding variables, such as circadian disruption and reduced activity levels, improving data quality and experimental reproducibility. At the same time, continuous monitoring supports the principles of refinement promoted by Directive 2010/63/EU and aligns with the recommendations outlined in the ARRIVE 2.0 guidelines.
The findings represent another step toward a future where digital technologies and artificial intelligence contribute to better welfare outcomes, more robust scientific data, and more informed decision-making in laboratory animal research.
Corporate
Italia
Australia/New Zealand
France
Germany
Japan
USA
Canada
UK
China
Switzerland