Estimating risk probabilities for sickness from behavioural patterns to identify health challenges in dairy cows with multivariate cumulative sum control charts.

Dairy cattle housing is characterised by increasing herd sizes and the need for assisting technical tools to monitor the cows' health. This study investigated the combination of logistic regression models with multivariate cumulative sum (MCUSUM) control charts in healthmonitoring of dairy cattle. Sensor information of 618 cows with 791 lactations (138 438 cow days), nine behavioural variables were included as parts of the behavioural patterns: physical activity ("neck activity", "leg activity", "walking duration"), resting ("lying duration", "standing duration", "transitions from lying to standing") and feeding ("feeding duration", "rumination duration", "inactivity duration") behaviour. For each of these behavioural patterns, a logistic regression model with the health status (sick vs not sick) as a dependent variable was designed after a variable selection (herd level) based on the herd dataset with 618 cows (618 lactations; 115 547 cow days), which included the variables of each behaviour pattern and the stage of lactation nested in the number of lactations as explanatory variables. The explanatory variables were added stepwise to the model, with the final model being selected with respect to the lowest values of Akaike's and Bayes' information criteria. Each model was then applied to a dataset with 173 cows (22 891 cow days) at cow level, resulting in individual daily risk probabilities for getting sick. Thus, risk probabilities of each behavioural pattern were estimated and included in the MCUSUM control charts to identify cows at risk of disease. The performance of the MCUSUM control charts was cross-validated to identify the best fitting reference value k and the threshold value h. Alerts given within 5 days prior to diagnosis were counted as detected sicknesses. The performance resulted in a block sensitivity of 70.9-81.4%, specificity of 87.9-94.2% and a false-positive rate of 5.8-12.1%. The performance was confirmed while testing the entire algorithm resulting in a mean area under the receiver operating characteristics curve of 0.89. Calculating precision and the F1-score resulted in a precision of 49.0-60.9% (training: 48.8-63.5%) and an F1-score of 61.1-65.7% in testing (training: 61.0-67.0%). The precision-recall curve (PRC) was derived from precision and recall with an area under the PRC of 0.70 in training and testing. In summary, the present study was able to develop an algorithm showing good classification potential for the online monitoring of sickness behaviour.

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

The present observational study investigated the application of multivariate cumulative sum (MCUSUM) 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.