A time-series approach for clustering farms based on slaughterhouse health aberration data

Abstract

A large amount of data is collected routinely in meat inspection in pig slaughterhouses. A time series clustering approach is presented and applied that groups farms based on similar statistical characteristics of meat inspection data over time. A three step characteristic-based clustering approach was used from the idea that the data contain more info than the incidence figures. A stratified subset containing 511,645 pigs was derived as a study set from 3.5 years of meat inspection data. The monthly averages of incidence of pleuritis and of pneumonia of 44 Dutch farms (delivering 5149 batches to 2 pig slaughterhouses) were subjected to 1) derivation of farm level data characteristics 2) factor analysis and 3) clustering into groups of farms. The characteristic-based clustering was able to cluster farms for both lung aberrations. Three groups of data characteristics were informative, describing incidence, time pattern and degree of autocorrelation. The consistency of clustering similar farms was confirmed by repetition of the analysis in a larger dataset. The robustness of the clustering was tested on a substantially extended dataset. This confirmed the earlier results, three data distribution aspects make up the majority of distinction between groups of farms and in these groups (clusters) the majority of the farms was allocated comparable to the earlier allocation (75% and 62% for pleuritis and pneumonia, respectively). The difference between pleuritis and pneumonia in their seasonal dependency was confirmed, supporting the biological relevance of the clustering. Comparison of the identified clusters of statistically comparable farms can be used to detect farm level risk factors causing the health aberrations beyond comparison on disease incidence and trend alone.

Agricultural Holdings and Slaughterhouses’ Impact on Patterns of Pathological Findings Observed during Post-Mortem Meat Inspection

“…By contrast, in Styria, where most pigs were slaughtered, the agricultural holding (R 2 = 0.33) explained more variability of findings in Category 1 than the slaughterhouses (R 2 = 0.26), although the slaughterhouse contribution was still quite considerable. These findings confirm observations from previous studies highlighting both the value of slaughterhouse data for feedback systems to farmers [ 31 ] as well as the need for improved recording procedures in the slaughterhouses as these introduce considerable variability, which may hamper the use of meat inspection data for a reliable feedback system to farmers [ 48 ].…”

“…Among the major pig-producing federal states, pleuritis and pneumonia were more frequently reported in Styria, whereas foreign body in the lung was much more frequently registered in Upper and Lower Austria. This might reflect a lack of standardisation in the current meat inspection system [ 48 ] or may be a result of the two different types of software implemented for documenting pathological findings in meat inspection in Austria [ 49 ].…”

“…Overall, most variance in the prevalence can be explained by the agricultural holding (average R 2 = 0.61), which may reflect differences in the type of pigs sent for slaughter, prevalence of specific diseases, farm management, type of housing, part or full-time occupation of farmers, size of holding and overall biosecurity level, which could be beneficial to herd health [ 11 , 45 , 46 , 47 ]. A clustering approach using meat inspection data from two Dutch slaughterhouses previously highlighted differences between groups of farms by lung abnormalities on the basis of data characteristics involving incidence, time pattern and degree of autocorrelation [ 48 ].…”

Development of a Porcine Slaughterhouse Kidney Perfusion Model

“…Furthermore, the conditions of an experiment with slaughterhouse organs are less controlled than in a laboratory setting. Many reports describe variability in bacterial load and underlying deceases [24][25][26]. The work environment is obviously different from a laboratory setting, since part of the preparation and retrieval of the kidneys takes place in a food processing plant.…”

KPIs-Based Clustering and Visualization of HPC Jobs: A Feature Reduction Approach

“…The set of features used (trend, seasonality, periodicity, serial correlation, skewness, kurtosis, chaos, nonlinearity, and selfsimilarity) achieved high accuracy clusters with benchmark time series datasets. In [62] the research work focused on farming, clustering farms based on similar time series statistical features of meat inspection in pig slaughterhouses. This information was useful to detect farm level risk factors causing health problems based on disease incidence and trend.…”