Non-Contact Activity Monitoring Using a Multi-Axial Inertial Measurement Unit in Animal Husbandry.

Area of Science:

• Animal behavior monitoring
• Biomedical engineering
• Sensor technology

Background:

• Accurate monitoring of animal physical activity is crucial for health evaluation and welfare.
• Traditional methods can be invasive or labor-intensive.
• Developing non-contact, non-invasive monitoring solutions is essential.

Purpose of the Study:

• To present a novel non-contact, non-invasive method for monitoring physical activity in animals.
• To classify different activity states including resting, stationary activity, and locomotion.
• To validate the method's accuracy and continuous monitoring capabilities.

Main Methods:

• Utilized a multi-axial inertial measurement unit (IMU) to measure structural vibrations.
• Mounted the IMU on the underside of a mouse husbandry cage floor.
• Collected 27.67 hours of IMU data and ground truth labels from camera recordings.
• Trained a classification model using accelerometer and angular rate data.

Main Results:

• Achieved over 90% accuracy in classifying activity versus non-activity.
• Successfully monitored activity continuously over 24 hours, revealing nocturnal behavior patterns.
• Demonstrated the ability to differentiate between resting, stationary activity, and locomotion.

Conclusions:

• The developed IMU-based method offers a powerful tool for non-contact, non-invasive physical activity assessment.
• This technology enables automatic health evaluation and optimization of animal care workflows.
• The findings contribute to improved animal wellbeing through enhanced monitoring capabilities.