Variable selection for monitoring sickness behavior in lactating dairy cattle with the application of control charts

Abstract

The present observational study investigated the application of multivariate cumulative sum (MCU-SUM) control charts by including variables selected by principal component analysis and partial least squares (PLS) regression to identify sickness behavior in dairy cattle. Therefore, sensor information (24 variables) was collected from 480 milking cows on a German dairy farm between September 2018 and December 2019. These variables were gathered in potentially different scenarios on farm. In total, data from 749 animals were available for evaluation. Variables were chosen based on the information of 499 cows (62 healthy; 437 sick) with 93,598 observations. The available diagnoses were collected together to form 1,025 sickness events. Hence, the different numbers of selected variables were included into the MCUSUM control charts. The performance of the MCUSUM control charts was evaluated by a 10-fold cross validation; hence, 90% of the original data set (749 cows) represented the training data, and the remaining 10% was used to test the training results. On average, the 10 training data sets included 124,871 observations with 1,392 sickness events, and the 10 testing data sets included, on average, 13,704 observations with 153 sickness events. The MCUSUM generated from the variables selected by principal component analysis showed comparable results in training and testing in all scenarios; therefore, 70.0 to 97.4% of the sickness events were detected. The false-positive rates ranged from 8.5 to 29.6%, and thus they created at least 2.6 false-positive alerts per day in testing. The variables selected by the PLS regression approach showed comparable sickness detection rates (70.0-99.9%) as well as false-positive rates (8.2-62.8%) in most scenarios. The best performing scenario produced 2.5 false-positive alerts in testing. Summarizing, both approaches showed potential for practical implementation; however, the PLS variable selection approach showed fewer false positives. Therefore, the PLS regression approach could generate a more reliable sickness detection algorithm, if combined with MCUSUM control charts, and considered for practical implementation.

Feeding behavior and activity levels are associated with recovery status in dairy calves treated with antimicrobials for Bovine Respiratory Disease

“…We suggest that there is a value to using behaviors collectively to indicate early onset of disease and recovery status in calves. Machine learning techniques using a precision technology device have previously been used to detect BRD in cattle 16 , 17 , 22 , 38 and are a promising area for future research. Collectively, we suggest that there is potential to detect recovery status of dairy calves using sickness behaviors, where relative changes in drinking speed, relative changes in calf starter intake, and relative changes in lying time might be especially useful behaviors.…”

“…In dairy calves, the most sensitive indicator of BRD was lying time, when using a random forest algorithm and rolling averages in lying time as machine learning techniques 16 . Depressed feeding behavior and reduced activity levels were also observed a few days before diagnosis of metabolic diseases in lactating dairy cattle 17 . Similarly, decreased feed intake 18 – 20 , increased lying times 20 and less lying bouts 21 were associated with BRD status when compared to healthy calves.…”

A randomized clinical trial evaluating the effects of administration of acidogenic boluses at dry-off on rumination and activity behavior in the 14 subsequent days

“…In our study, cows in the TRT group spent less time ruminating on the first day after dry-off but increased their TDR in the following 3 d to the same levels as cows in the CON group with the same rumination behavior pattern observed in both groups for the remainder of the study period. Adverse health and welfare events lead to changes in different behavioral patterns of dairy cattle, such as resting, feeding, rumination, activity, and even socialization behaviors ( Dittrich et al, 2021 ). Due to its biological importance, changes in rumination have long been associated with impaired health in dairy cows ( Radostits et al, 2007 ).…”

Do device that simulate methane capture equipment, as a mask located in the head of dairy cows affect their welfare?

“…Behavior (eating, ruminating and inactive) data were collected by “CowScout” sensor systems (GEA Farm Technologies, Aktiengesellschaft, Bönen, Germany) which were attached to the neck collar [ 13 ]. Data were recorded daily from day −6 to day 28, in relation to the moment of placing the methane capture device on the animals.…”