Prediction of the start of parturition in mares using camera image analysis.

Prediction of the start of labor in pregnant mares was studied by analyzing changes in their movement behavior and body surface temperature using camera image analysis, showing increased movement and temperature prior to parturition.

Objective and Study Design

• The study aimed to predict the onset of labor (parturition) in pregnant Thoroughbred mares by detecting changes in their behavior and body surface temperature (BST).
• 53 pregnant mares were individually housed in box stalls and monitored continuously using a digital camera system with both thermal and visible light cameras.

Methodology: Image and Thermal Analysis

• Image processing software was used to automatically identify the centroid (central point) of each mare from the visible camera footage, determining x and y coordinates.
• Trajectory analysis software utilized these coordinates to calculate the distance moved by each mare over time.
• Body surface temperature was measured with infrared thermography, providing average BST readings.
• To account for environmental effects, the difference between BST and ambient temperature was used rather than absolute BST values.

Key Findings on Movement and Temperature

• On the day labor began (start of the expulsion phase, BEP), mares showed a significant increase in movement compared to their activity two days before BEP.
• The peak movement increase was observed 1 to 2 hours before the start of labor.
• BST showed a tendency to increase within 6 hours before BEP, with a statistically significant rise in the last hour prior to labor (1-0 hours before BEP).

Significance of Combined Behavioral and Temperature Data

• Overall, 96.2% of the mares exhibited either increased activity or raised BST before labor.
• This indicates that monitoring multiple parameters (movement and temperature) improves the ability to predict the onset of parturition.
• The study suggests that camera-based automated analysis systems can provide valuable, non-invasive prediction tools for managing pregnant mares and preparing for foaling.