An Information-Theoretic Approach to Detect the Associations of GPS-Tracked Heifers in Pasture.

Sensor technologies, such as the Global Navigation Satellite System (GNSS), produce huge amounts of data by tracking animal locations with high temporal resolution. Due to this high resolution, all animals show at least some co-occurrences, and the pure presence or absence of co-occurrences is not satisfactory for social network construction. Further, tracked animal contacts contain noise due to measurement errors or random co-occurrences. To identify significant associations, null models are commonly used, but the determination of an appropriate null model for GNSS data by maintaining the autocorrelation of tracks is challenging, and the construction is time and memory consuming. Bioinformaticians encounter phylogenetic background and random noise on sequencing data. They estimate this noise directly on the data by using the average product correction procedure, a method applied to information-theoretic measures. Using Global Positioning System (GPS) data of heifers in a pasture, we performed a proof of concept that this approach can be transferred to animal science for social network construction. The approach outputs stable results for up to 30% missing data points, and the predicted associations were in line with those of the null models. The effect of different distance thresholds for contact definition was marginal, but animal activity strongly affected the network structure.

Detecting Animal Contacts-A Deep Learning-Based Pig Detection and Tracking Approach for the Quantification of Social Contacts.

The identification of social interactions is of fundamental importance for animal behavioral studies, addressing numerous problems like investigating the influence of social hierarchical structures or the drivers of agonistic behavioral disorders. However, the majority of previous studies often rely on manual determination of the number and types of social encounters by direct observation which requires a large amount of personnel and economical efforts. To overcome this limitation and increase research efficiency and, thus, contribute to animal welfare in the long term, we propose in this study a framework for the automated identification of social contacts. In this framework, we apply a convolutional neural network (CNN) to detect the location and orientation of pigs within a video and track their movement trajectories over a period of time using a Kalman filter (KF) algorithm. Based on the tracking information, we automatically identify social contacts in the form of head-head and head-tail contacts. Moreover, by using the individual animal IDs, we construct a network of social contacts as the final output. We evaluated the performance of our framework based on two distinct test sets for pig detection and tracking. Consequently, we achieved a Sensitivity, Precision, and F1-score of 94.2%, 95.4%, and 95.1%, respectively, and a MOTA score of 94.4%. The findings of this study demonstrate the effectiveness of our keypoint-based tracking-by-detection strategy and can be applied to enhance animal monitoring systems.

Panoptic Segmentation of Individual Pigs for Posture Recognition.

Behavioural research of pigs can be greatly simplified if automatic recognition systems are used. Systems based on computer vision in particular have the advantage that they allow an evaluation without affecting the normal behaviour of the animals. In recent years, methods based on deep learning have been introduced and have shown excellent results. Object and keypoint detector have frequently been used to detect individual animals. Despite promising results, bounding boxes and sparse keypoints do not trace the contours of the animals, resulting in a lot of information being lost. Therefore, this paper follows the relatively new approach of panoptic segmentation and aims at the pixel accurate segmentation of individual pigs. A framework consisting of a neural network for semantic segmentation as well as different network heads and postprocessing methods will be discussed. The method was tested on a data set of 1000 hand-labeled images created specifically for this experiment and achieves detection rates of around 95% (F1 score) despite disturbances such as occlusions and dirty lenses.

Using Acceleration Data to Automatically Detect the Onset of Farrowing in Sows.

The aim of the present study was to automatically predict the onset of farrowing in crate-confined sows. (1) Background: Automatic tools are appropriate to support animal surveillance under practical farming conditions. (2) Methods: In three batches, sows in one farrowing compartment of the Futterkamp research farm were equipped with an ear sensor to sample acceleration. As a reference video, recordings of the sows were used. A classical CUSUM chart using different acceleration indices of various distribution characteristics with several scenarios were compared. (3) Results: The increase of activity mainly due to nest building behavior before the onset of farrowing could be detected with the sow individual CUSUM chart. The best performance required a statistical distribution characteristic that represented fluctuations in the signal (for example, 1st variation) combined with a transformation of this parameter by cumulating differences in the signal within certain time periods from one day to another. With this transformed signal, farrowing sows could reliably be detected. For 100% or 85% of the sows, an alarm was given within 48 or 12 h before the onset of farrowing. (4) Conclusions: Acceleration measurements in the ear of a sow are suitable for detecting the onset of farrowing in individually housed sows in commercial farrowing crates.

Towards automatic farrowing monitoring-A Noisy Student approach for improving detection performance of newborn piglets.

Nowadays, video monitoring of farrowing and automatic video evaluation using Deep Learning have become increasingly important in farm animal science research and open up new possibilities for addressing specific research questions like the determination of husbandry relevant indicators. A robust detection performance of newborn piglets is essential for reliably monitoring the farrowing process and to access important information about the welfare status of the sow and piglets. Although object detection algorithms are increasingly being used in various scenarios in the field of livestock farming, their usability for detecting newborn piglets has so far been limited. Challenges such as frequent animal occlusions, high overlapping rates or strong heterogeneous animal postures increase the complexity and place new demands on the detection model. Typically, new data is manually annotated to improve model performance, but the annotation effort is expensive and time-consuming. To address this problem, we propose a Noisy Student approach to automatically generate annotation information and train an improved piglet detection model. By using a teacher-student model relationship we transform the image structure and generate pseudo-labels for the object classes piglet and tail. As a result, we improve the initial detection performance of the teacher model from 0.561, 0.838, 0.672 to 0.901, 0.944, 0.922 for the performance metrics Recall, Precision and F1-score, respectively. The results of this study can be used in two ways. Firstly, the results contribute directly to the improvement of piglet detection in the context of birth monitoring systems and the evaluation of the farrowing progress. Secondly, the approach presented can be transferred to other research questions and species, thereby reducing the problem of cost-intensive annotation processes and increase training efficiency. In addition, we provide a unique dataset for the detection and evaluation of newborn piglets and sow body parts to support researchers in the task of monitoring the farrowing process.

Detecting tail biters by monitoring pig screams in weaning pigs.

Early identification of tail biting and intervention are necessary to reduce tail lesions and their impact on animal health and welfare. Removal of biters has become an effective intervention strategy, but finding them can be difficult and time-consuming. The aim of this study was to investigate whether tail biting and, in particular, individual biters could be identified by detecting pig screams in audio recordings. The study included 288 undocked weaner pigs housed in six pens in two batches. Once a tail biter (n = 7) was identified by visual inspection in the stable and removed by the farm staff, the previous days of video and audio recordings were analyzed for pig screams (sudden increase in loudness with frequencies above 1 kHz) and tail biting events until no biting before the removal was observed anymore. In total, 2893 screams were detected in four pens where tail biting occurred. Of these screams, 52.9% were caused by tail biting in the observed pen, 25.6% originated from other pens, 8.8% were not assignable, and 12.7% occurred due to other reasons. In case of a tail biting event, screams were assigned individually to biter and victim pigs. Based on the audio analysis, biters were identified between one and nine days prior to their removal from the pen after visual inspection. Screams were detected earlier than the increase in hanging tails and could therefore be favored as an early warning indicator. Analyzing animal vocalization has potential for monitoring and early detection of tail biting events. In combination with individual marks and automatic analysis algorithms, biters could be identified and tail biting efficiently reduced. In this way, biters can be removed earlier to increase animal health and welfare.

Implementation of multivariate cumulative sum control charts in mastitis and lameness monitoring.

This study analyzed the methodology and applicability of multivariate cumulative sum (MCUSUM) charts for early mastitis and lameness detection. Data used were recorded on the Karkendamm dairy research farm, Germany, between August 2008 and December 2010. Data of 328 and 315 cows in their first 200 d in milk were analyzed for mastitis and lameness detection, respectively. Mastitis as well as lameness was specified according to veterinary treatments. Both diseases were defined as disease blocks. Different disease definitions for mastitis and lameness (2 for mastitis and 3 for lameness) varied solely in the sequence length of the blocks. Only the days before the treatment were included in the disease blocks. Milk electrical conductivity, milk yield, and feeding patterns (feed intake, number of trough visits, and feeding time) were used for the recognition of mastitis. Pedometer activity and feeding patterns were used for lameness detection. To exclude biological trends and obtain independent observations, the values of each input variable were either preprocessed by wavelet filters or a multivariate vector autoregressive model. The residuals generated between the observed and filtered or observed and forecast values, respectively, were then transferred to a classic or self-starting MCUSUM chart. The combination of the 2 preprocessing methods with each of the 2 MCUSUM sum charts resulted in 4 combined monitoring systems. For mastitis as well as lameness detection requiring a block sensitivity of at least 70%, all 4 of the combined monitoring systems used revealed similar results within each of the disease definitions. Specificities of 73 to 80% and error rates of 99.6% were achieved for mastitis. The results for lameness showed that the definitions used obtained specificities of up to 81% and error rates of 99.1%. The results indicate that the monitoring systems with these study characteristics have appealing features for mastitis and lameness detection. However, they are not yet directly applicable for practical implementations.

Principal component analysis for the early detection of mastitis and lameness in dairy cows.

This investigation analysed the applicability of principal component analysis (PCA), a latent variable method, for the early detection of mastitis and lameness. Data used were recorded on the Karkendamm dairy research farm between August 2008 and December 2010. For mastitis and lameness detection, data of 338 and 315 cows in their first 200 d in milk were analysed, respectively. Mastitis as well as lameness were specified according to veterinary treatments. Diseases were defined as disease blocks. The different definitions used (two for mastitis, three for lameness) varied solely in the sequence length of the blocks. Only the days before the treatment were included in the blocks. Milk electrical conductivity, milk yield and feeding patterns (feed intake, number of feeding visits and time at the trough) were used for recognition of mastitis. Pedometer activity and feeding patterns were utilised for lameness detection. To develop and verify the PCA model, the mastitis and the lameness datasets were divided into training and test datasets. PCA extracted uncorrelated principle components (PC) by linear transformations of the raw data so that the first few PCs captured most of the variations in the original dataset. For process monitoring and disease detection, these resulting PCs were applied to the Hotelling's T 2 chart and to the residual control chart. The results show that block sensitivity of mastitis detection ranged from 77·4 to 83·3%, whilst specificity was around 76·7%. The error rates were around 98·9%. For lameness detection, the block sensitivity ranged from 73·8 to 87·8% while the obtained specificities were between 54·8 and 61·9%. The error rates varied from 87·8 to 89·2%. In conclusion, PCA seems to be not yet transferable into practical usage. Results could probably be improved if different traits and more informative sensor data are included in the analysis.

Characteristics of thermal images of the mammary gland and of performance in sows differing in health status and parity.

Precision livestock farming can combine sensors and complex data to provide a simple score of meaningful productivity, pig welfare, and farm sustainability, which are the main drivers of modern pig production. Examples include using infrared thermography to monitor the temperature of sows to detect the early stages of the disease. To take account of these drivers, we assigned 697 hybrid (BHZP db. Viktoria) sows to four parity groups. In addition, by pooling clinical findings from every sow and their piglets, sows were classified into three groups for the annotation: healthy, clinically suspicious, and diseased. Besides, the udder was thermographed, and performance data were documented. Results showed that the piglets of diseased sows with eighth or higher parity had the lowest daily weight gain [healthy; 192 g ± 31.2, clinically suspicious; 191 g ± 31.3, diseased; 148 g ± 50.3 (p < 0.05)] and the highest number of stillborn piglets (healthy; 2.2 ± 2.39, clinically suspicious; 2.0 ± 1.62, diseased; 3.91 ± 4.93). Moreover, all diseased sows showed higher maximal skin temperatures by infrared thermography of the udder (p < 0.05). Thus, thermography coupled with Artificial Intelligence (AI) systems can help identify and orient the diagnosis of symptomatic animals to prompt adequate reaction at the earliest time.

Effects of Different Housing Systems during Suckling and Rearing Period on Skin and Tail Lesions, Tail Losses and Performance of Growing and Finishing Pigs.

Feasible alternatives to stressful weaning and tail-docking are needed to inhibit tail biting. Therefore, we investigated the effects of housing systems for 1106 pigs that were weaned from: (1) conventional farrowing crates (FC), (2) free-farrowing pens (FF), or (3) group housing of lactating sows (GH) into (1) conventional rearing pens (Conv) or (2) piglets remained in their farrowing pens for rearing (Reaf). Tails were docked or left undocked batchwise. All pigs were regrouped for the fattening period. Pigs were scored for skin lesions, tail lesions and losses. After weaning, Conv-GH pigs had significantly less skin lesions than Conv-FC and Conv-FF pigs. After regrouping for fattening, Reaf-GH pigs had significantly less skin lesions than Conv pigs, Reaf-FC and Reaf-FF. The frequency of tail lesions of undocked Conv pigs peaked in week 4 (66.8%). Two weeks later, Reaf undocked pigs reached their maximum (36.2%). At the end of fattening, 99.3% of undocked Conv pigs and 43.1% of undocked Reaf pigs lost parts of their tail. In conclusion, the co-mingling of piglets during suckling reduced the incidence of skin lesions. Rearing in the farrowing pen significantly reduced the incidence of tail lesions and losses for undocked pigs. No housing system negatively affected the performance.

Classification of Pigs with Tail Lesions from Different Farrowing and Rearing Systems during Rearing and Fattening Period.

The aim of the present study was to classify and characterise pigs with tail lesions using a combined parameter based on the frequency and duration of tail lesions and to find out whether biologically relevant groups could be separated by cluster analysis. Pigs (n = 677, 50% docked, 50% undocked) from three farrowing systems, as follows: (1) Conventional farrowing crate (FC), (2) free farrowing (FF), and (3) a group housing lactating sows (GH), were divided into two rearing systems as follows: (1) A conventional system (CONV) and (2) a wean-to-finish (W-F) system. Within 18 assessment weeks, starting after weaning, animal tail lesions were recorded individually. The animals were characterised into five lesion groups, as follows: (I) No lesions to (V) many long lasting lesions. The separability of the predefined lesion groups was checked by an animal individual lesion parameter. By using a k-means cluster analysis, it was shown that the docking status was the mainly affected parameter on the tail lesions. The separation of the groups only succeeded for the most distinct groups, I and V. The high impact of the docking status and the reduction of tail lesions by more space allowance was shown. More characterising information for the individual pigs would improve the separability of the lesion groups.

Analysis of risk factors for infections with gastrointestinal nematodes, Eimeria spp. and lungworms in German organic sheep farms.

The aim of this study was to determine the risk factors influencing the occurrence of parasitic infections in organic sheep farms in Germany. Therefore, 635 pooled faecal samples from sheep kept on 20 organic farms were collected and examined by standard parasitological analyses for gastrointestinal nematodes (GINs), Eimeria species (spp.) and liver flukes (Fasciola hepatica). Additionally, 128 double-pooled samples were analysed for lungworm larvae. In 60.5% of all samples, parasite stadiums were detected, and 38.3% of the double-pooled samples were lungworm-positive. Production period, months and year of sampling had significant effects on infections with GINs (p < 0.05). The prevalence of GIN infection was lowest in 'dairy'(40.0%) when compared with'meat'sheep (65.4%). The odds of being infected with Eimeria spp. was influenced by the month (p < 0.05). The number of ewes on a farm, the primary purpose or the grazing area showed no significant effects. Infections with lungworms occurred in tendency more often 'after' lambing period.

Efficient interruption of infection chains by targeted removal of central holdings in an animal trade network.

Centrality parameters in animal trade networks typically have right-skewed distributions, implying that these networks are highly resistant against the random removal of holdings, but vulnerable to the targeted removal of the most central holdings. In the present study, we analysed the structural changes of an animal trade network topology based on the targeted removal of holdings using specific centrality parameters in comparison to the random removal of holdings. Three different time periods were analysed: the three-year network, the yearly and the monthly networks. The aim of this study was to identify appropriate measures for the targeted removal, which lead to a rapid fragmentation of the network. Furthermore, the optimal combination of the removal of three holdings regardless of their centrality was identified. The results showed that centrality parameters based on ingoing trade contacts, e.g. in-degree, ingoing infection chain and ingoing closeness, were not suitable for a rapid fragmentation in all three time periods. More efficient was the removal based on parameters considering the outgoing trade contacts. In all networks, a maximum percentage of 7.0% (on average 5.2%) of the holdings had to be removed to reduce the size of the largest component by more than 75%. The smallest difference from the optimal combination for all three time periods was obtained by the removal based on out-degree with on average 1.4% removed holdings, followed by outgoing infection chain and outgoing closeness. The targeted removal using the betweenness centrality differed the most from the optimal combination in comparison to the other parameters which consider the outgoing trade contacts. Due to the pyramidal structure and the directed nature of the pork supply chain the most efficient interruption of the infection chain for all three time periods was obtained by using the targeted removal based on out-degree.

Static network analysis of a pork supply chain in Northern Germany-Characterisation of the potential spread of infectious diseases via animal movements.

Transport of live animals is a major risk factor in the spread of infectious diseases between holdings. The present study analysed the pork supply chain of a producer community in Northern Germany. The structure of trade networks can be characterised by carrying out a network analysis. To identify holdings with a central position in this directed network of pig production, several parameters describing these properties were measured (in-degree, out-degree, ingoing and outgoing infection chain, betweenness centrality and ingoing and outgoing closeness centrality). To obtain the importance of the different holding types (multiplier, farrowing farms, finishing farms and farrow-to-finishing farms) within the pyramidal structure of the pork supply chain, centrality parameters were calculated for the entire network as well as for the individual holding types. Using these centrality parameters, two types of holdings could be identified. In the network studied, finishing and farrow-to-finishing farms were more likely to be infected due to the high number of ingoing trade contacts. Due to the high number of outgoing trade contacts multipliers and farrowing farms had an increased risk to spread a disease to other holdings. However, the results of the centrality parameters degree and infection chain were not always consistent, such that the indirect trade contacts should be taken into consideration to understand the real importance of a holding in spreading or contracting an infection. Furthermore, all calculated parameters showed a highly right-skewed distribution. Networks with such a degree distribution are considered to be highly resistant concerning the random removal of nodes. But by strategic removal of the most central holdings, e.g. by trade restrictions or selective vaccination or culling, the network structure can be changed efficiently and thus decompose into fragments. Such a fragmentation of the trade networks is of particular importance from an epidemiological perspective.

The additional costs of segregated transport to slaughter to decrease Salmonella prevalence in pork--a simulation study.

A stochastic simulation model was used to assess the additional costs related to segregated transport to slaughter. This control measure was assumed to be implemented within a producers' association to decrease Salmonella prevalence in pork. Calculations were based on the additional shipments caused by the separate transport of low- and high-prevalence herds and on the additional transport distance caused by changed routing. The results showed that there is not necessarily a considerable increase in the number of shipments due to herd status separation for transport. The percentage of shipments changed due to segregated transport varied between 43% and 69% depending on the threshold prevalence. The additional costs per slaughtered pig varied between 0.07€/pig and 0.58€/pig under the given assumptions. Costs were governed by the percentage of changed shipments and the additional distance of a changed shipment. Due to the fact that the percentage of changed shipments is related to the distribution of herd prevalence within the producers' association, there is no cost-effective threshold in general. Different producers' associations incur different costs caused by segregated transport to slaughter at the same threshold prevalence. The current study supports producers' associations in evaluating the additional costs of segregated transport for their members.

Assessing individual sow risk factors for coliform mastitis: a case-control study.

In order to investigate sow-specific risk factors associated with coliform mastitis, a case-control study was performed over the course of 28 months. Data of three farms were collected under production conditions. Sows suffering from coliform mastitis after farrowing served as cases, and healthy half- or full-sib sows from the same farm served as controls. Individual sow characteristics and the seasonal influence were analysed by conditional logistic regression. The final multivariate model identified four risk factors: the risk of suffering from coliform mastitis increased with a higher number of piglets born alive and stillborn piglets. Gilts had an increased risk for the disease, and birth intervention was also associated with a higher prevalence of mastitis. Birth induction and season had no significant influence on the occurrence of coliform mastitis. The time during and soon after farrowing is a very sensitive period in pig production demanding great attention by the farmer. With respect to the economic losses, monitoring of potentially endangered sows as well as detailed documentation and selection of disease cases are of particular importance when coping with coliform mastitis.

Comparison of virulence gene profiles of Escherichia coli isolates from sows with coliform mastitis and healthy sows.

Coliform mastitis (CM) is not only a serious economical and animal welfare touching problem in dairy cattle, but also in sows after farrowing. Due to this disease, the essential adequate supply with colostrum for the growth and the health of the piglets is not ensured. Besides other influencing factors, Escherichia (E.) coli is of great importance as a causative agent of this multifactorial disease. In this study, E. coli isolates from milk samples of healthy and CM-affected sows were examined for the presence of virulence genes associated with extraintestinal E. coli strains, enterotoxigenic E. coli and other pathogenic E. coli. The isolated E. coli harbored mainly virulence genes of extraintestinal E. coli strains (especially fimC, ompA, traT, hra, kpsMTII, iroN). The virulence gene spectrum for both samples from CM-affected and healthy sows did not differ significantly. Particular virulence gene profiles of E. coli isolates from diseased sows were not detected. This study provides novel insights into the role of E. coli in association with mastitis in sows since it is the first time E. coli isolates from CM-affected sows' milk were analysed for virulence genes. Because there were no differences in the prevalence of E. coli and their virulence-associated genes between healthy and diseased sows, other causative factors seem to have greater influence on the pathogenesis of porcine CM.

Consideration of different outbreak conditions in the evaluation of preventive culling and emergency vaccination to control foot and mouth disease epidemics.

In recent foot and mouth disease outbreaks, many healthy animals have been culled to prevent disease transmission. Emergency vaccination is discussed as an alternative to culling of unaffected animals. A spatial and temporal Monte-Carlo simulation model was used to compare preventive culling and emergency vaccination. Different outbreaks are described using additional influence factors such as airborne spread, farm density, type of index-case farm and delay until establishment of the control strategies. The fewest farms were infected establishing a combined strategy including a 1 km preventive culling and 1-10 km emergency vaccination zone around each outbreak farm. Taking the number of culled and vaccinated farms into account, vaccination around the first diagnosed farm combined with the baseline strategy (culling of outbreak farms, protection and surveillance zone, contact tracing) is to be preferred. In the present study, emergency vaccination was an effective control strategy especially in densely populated regions.