Predicting the onset of left displaced abomasum using behavioral changes measured by small 3D acceleration sensors attached to a cow collar.

The present study investigated whether behavioral changes predict the onset of left displaced abomasum (LDA) using a small, multi-sensor device attached to a cow collar that records data in real time over a 24 hr period. Behavioral data were collected from 17 LDA cows that had undergone surgery and 23 healthy dairy cows. The total rumination time over 10 days before and after LDA surgery in cows, was significantly shorter from 8 days before to 3 days after surgery, peaking at 216.4 ± 96.8 min on day before surgery than that in the control group (group: P<0.001). The total rumination time of LDA cows was significantly shorter than that of the control group at each time point (interaction effect: P<0.001). The diagnostic ability to predict the onset of LDA based on the cutoff value (≤360 min) of total rumination time 6 days before LDA surgery calculated using the Yohden index showed a significantly high AUC of 0.887 (P<0.0001) with high sensitivity (84.6%) and specificity (88.9%). In other words, it was suggested that the onset of LDA can be predicted by total rumination time from 6 days before surgery. An Animal Health Monitoring System (AHMS) not only warns of the onset of cattle diseases, such as LDA, several days in advance, but may also predict the onset of diseases up to 6 days in advance based on shortened the total rumination times, thereby contributing to the efficiency of work for farmers and veterinarians.