Reticuloruminal Motility Monitoring for the Prediction of Peripartal Hypocalcemia in Cattle

Area of Science:

• Veterinary Medicine
• Animal Science
• Dairy Production

Background:

• Hypocalcemia (low blood calcium) is a common metabolic disorder in dairy cows during the transition period.
• Accurate prediction of hypocalcemia risk allows for tailored preventive strategies and interventions.

Purpose of the Study:

• To measure total (tCa) and ionized calcium (iCa) levels during the peripartal period.
• To record rumen motility and temperature patterns using reticuloruminal bolus sensors.
• To assess the association between pre-calving rumen patterns and calving-associated calcium levels.

Main Methods:

• Monitored 89 calvings (47 cows, 22 heifers) using reticuloruminal bolus sensors for 60 days pre- and postpartum.
• Recorded rumen motility, temperature, and analyzed total and ionized calcium concentrations.
• Evaluated a rumination time (RT) threshold and a deep learning model for hypocalcemia prediction.

Main Results:

• Cows with low tCa or iCa pre-parturition exhibited reduced rumination times.
• A rumination time threshold predicted hypocalcemia with 66.7% sensitivity and 96.8% specificity.
• A deep learning model achieved higher prediction accuracy (83.2% sensitivity, 98.2% specificity) than the RT threshold.

Conclusions:

• Multi-metric sensor analytics, particularly deep learning models, offer superior performance for early hypocalcemia risk detection in dairy cows.
• The deep learning approach demonstrates greater generalizability for improved early detection and management of hypocalcemia.