Evaluation of the tri-axial accelerometer to identify and predict parturition-related activities of Debouillet ewes in an intensive setting

摘要: Abstract Identifying and monitoring parturition of individual animals may help producers increase attentiveness, enabling early detection dystocia during parturition. Parturition events are marked by subtle behavioral changes often difficult to detect observation alone. The aim this study was determine the ability tri-axial accelerometer data accurately identify predict parturition-related behavior mature ewes in a pen setting. Tri-axial accelerometers recording at 12.5 Hz were placed on ear tags attached 13 Debouillet before Activity monitored 7 days prior lambing (d −7); day 0); post +7). Using random forest machine learning, visual observations used (i) seven mutually-exclusive behaviors; (ii) activity (active inactive behavior) based five metrics calculated using variation movements recorded accelerometer. accuracy predicted behaviors from an independent validation set 66.7 %, for 87.2 %. In addition activity, predictions evaluated d after 12 h six where actual time observed. No differences detected either or lambing. Four (P ≤ 0.002) higher than Values three highest