Walking kinematics of growing pigs associated with differences in musculoskeletal conformation, subjective gait score and osteochondrosis

A workflow for automatic, high precision livestock diagnostic screening of locomotor kinematics

Locomotor kinematics have been challenging inputs for automated diagnostic screening of livestock. Locomotion is a highly variable behavior, and influenced by subject characteristics (e.g., body mass, size, age, disease). We assemble a set of methods from different scientific disciplines, composing an automatic, high through-put workflow which can disentangle behavioral complexity and generate precise individual indicators of non-normal behavior for application in diagnostics and research. For this study, piglets (Sus domesticus) were filmed from lateral perspective during their first 10 h of life, an age at which maturation is quick and body mass and size have major consequences for survival. We then apply deep learning methods for point digitization, calculate joint angle profiles, and apply information-preserving transformations to retrieve a multivariate kinematic data set. We train probabilistic models to infer subject characteristics from kinematics. Model accuracy was validated for strides from piglets of normal birth weight (i.e., the category it was trained on), but the models infer the body mass and size of low birth weight (LBW) piglets (which were left out of training, out-of-sample inference) to be "normal." The age of some (but not all) low birth weight individuals was underestimated, indicating developmental delay. Such individuals could be identified automatically, inspected, and treated accordingly. This workflow has potential for automatic, precise screening in livestock management.

Modeling gait score of broiler chicken via production and behavioral data

Lameness in broilers may be associated with pain and is considered a major broiler production and welfare concern. Manual gait score assessment in commercial broiler houses is discrete, time-consuming, and laborious. As such, automatic methods for broiler gait score assessment are urgently needed. The objective of this study was to identify the relation of broiler gait score with several productions and behavioral metrics (bird BW, age, activity, and distribution), and establish three gait score prediction models for automatic gait score estimations in broiler farms with automatic weighing systems, camera systems, or both. Sixteen pens were used to rear Cobb 500 and Ross 708 broilers for eight and nine weeks, respectively (eight pens/strain, 12 birds/pen). The gait scores of all birds were assessed weekly by trained assessors following a six-point (0-5) scoring protocol from the third week. The pen's average BW was measured weekly. Top-view cameras were installed to continuously record videos of broilers in all 16 pens. Images were extracted from video clips (10 min/hour) during a 16-hour light period to determine the activity index and distribution index through image processing. The gait score was positively correlated with BW (R² = 0.97 for Cobb and R² = 0.96 for Ross), while negatively correlated with activity (R² = 0.78 for Cobb and R² = 0.73 for Ross). The three models showed high accuracies in predicting broiler gait score based on variables of BW, age, activity index, and distribution index (R² = 0.90-0.91, RMSE = 0.38-0.41). The findings of this study demonstrated the potential of estimating broiler gait score using bird BW, age, activity index, and distribution index. This information will assist in the development of automated gait score assessment systems in broiler production.

Image Analysis and Computer Vision Applications in Animal Sciences: An Overview

Computer Vision, Digital Image Processing, and Digital Image Analysis can be viewed as an amalgam of terms that very often are used to describe similar processes. Most of this confusion arises because these are interconnected fields that emerged with the development of digital image acquisition. Thus, there is a need to understand the connection between these fields, how a digital image is formed, and the differences regarding the many sensors available, each best suited for different applications. From the advent of the charge-coupled devices demarking the birth of digital imaging, the field has advanced quite fast. Sensors have evolved from grayscale to color with increasingly higher resolution and better performance. Also, many other sensors have appeared, such as infrared cameras, stereo imaging, time of flight sensors, satellite, and hyperspectral imaging. There are also images generated by other signals, such as sound (ultrasound scanners and sonars) and radiation (standard x-ray and computed tomography), which are widely used to produce medical images. In animal and veterinary sciences, these sensors have been used in many applications, mostly under experimental conditions and with just some applications yet developed on commercial farms. Such applications can range from the assessment of beef cuts composition to live animal identification, tracking, behavior monitoring, and measurement of phenotypes of interest, such as body weight, condition score, and gait. Computer vision systems (CVS) have the potential to be used in precision livestock farming and high-throughput phenotyping applications. We believe that the constant measurement of traits through CVS can reduce management costs and optimize decision-making in livestock operations, in addition to opening new possibilities in selective breeding. Applications of CSV are currently a growing research area and there are already commercial products available. However, there are still challenges that demand research for the successful development of autonomous solutions capable of delivering critical information. This review intends to present significant developments that have been made in CVS applications in animal and veterinary sciences and to highlight areas in which further research is still needed before full deployment of CVS in breeding programs and commercial farms.

Comparison of data analytics strategies in computer vision systems to predict pig body composition traits from 3D images

Computer vision systems (CVS) have been shown to be a powerful tool for the measurement of live pig body weight (BW) with no animal stress. With advances in precision farming, it is now possible to evaluate the growth performance of individual pigs more accurately. However, important traits such as muscle and fat deposition can still be evaluated only via ultrasound, computed tomography, or dual-energy x-ray absorptiometry. Therefore, the objectives of this study were: 1) to develop a CVS for prediction of live BW, muscle depth (MD), and back fat (BF) from top view 3D images of finishing pigs and 2) to compare the predictive ability of different approaches, such as traditional multiple linear regression, partial least squares, and machine learning techniques, including elastic networks, artificial neural networks, and deep learning (DL). A dataset containing over 12,000 images from 557 finishing pigs (average BW of 120 ± 12 kg) was split into training and testing sets using a 5-fold cross-validation (CV) technique so that 80% and 20% of the dataset were used for training and testing in each fold. Several image features, such as volume, area, length, widths, heights, polar image descriptors, and polar Fourier transforms, were extracted from the images and used as predictor variables in the different approaches evaluated. In addition, DL image encoders that take raw 3D images as input were also tested. This latter method achieved the best overall performance, with the lowest mean absolute scaled error (MASE) and root mean square error for all traits, and the highest predictive squared correlation (R2). The median predicted MASE achieved by this method was 2.69, 5.02, and 13.56, and R2 of 0.86, 0.50, and 0.45, for BW, MD, and BF, respectively. In conclusion, it was demonstrated that it is possible to successfully predict BW, MD, and BF via CVS on a fully automated setting using 3D images collected in farm conditions. Moreover, DL algorithms simplified and optimized the data analytics workflow, with raw 3D images used as direct inputs, without requiring prior image processing.

Alleged predisposing dietary factors fail to increase the incidence of osteochondrosis-like lesions in growing pigs at 14 and 24 wk of age

Early lesions of osteochondrosis (OC) are exhibited by regions of cartilage retention along the growth plate and articular cartilage. Progression of OC lesions may impair locomotion and necessitate euthanasia in adherence to animal welfare guides. Little is known about the role of nutrition in the initiation and early stages of OC. However, dietary components are commonly implicated as predisposing factors. In this study, diets were altered as an attempt to induce early stage OC lesions under controlled conditions. At 8 wk of age, 96 crossbred gilts (body weight [BW] = 17.4 ± 0.18 kg) were randomly assigned to one of four corn-soybean meal-based diets (four pens per diet, six pigs per pen) to assess diet effects on the number and volume of OC lesions in the distal femur. Diets included a non-pelleted control diet (Ctl); Ctl plus 20% glucose (Glc); the Ctl with increased concentrations of lysine, Ca, and P (+CaP); and the +CaP diet in a pelleted form (PEL). Femurs were collected from pigs euthanized at either 14-wk (Wk 14) or 24-wk (Wk 14) of age for assessments of OC lesions. Based on a mixed model analysis with pen as the experimental unit, dietary treatments did not affect final BW (129.3 ± 3.8 kg) or average daily gain (ADG) (1.00 ± 0.03 kg/d) over the trial. As expected, pigs fed PEL and Glc diets were more efficient (P < 0.05) in feed conversion compared with Ctl and +CaP. Using femurs as the experimental unit at Wk 14 (collected from two of the six pigs per pen), bone mineral content, determined by dual-energy x-ray absorptiometry scans, was greater (P < 0.05) in pigs fed +CaP and PEL than Ctl or Glc diets; however, only +CaP group differed (P < 0.05) at Wk 24 (collected from four pigs per pen). Computed tomography (CT) scans of femurs were reconstructed as three-dimensional images to allow detection of the number, volume, and surface area of lesions in distal growth plates. At Wk 14, pigs fed Ctl had fewer number of lesions (P < 0.05); however, no differences were detected among dietary treatments in lesion volume or lesion surface area. Pigs had fewer lesions at Wk 24 than Wk 14; however, differences were not detected among dietary treatments. At Wk 24, pigs fed Ctl diets had the greatest lesion volume among dietary treatments (P < 0.05). In conclusion, none of the pigs exhibited symptoms of lameness regardless of dietary treatment or OC lesion traits. Diet modifications due to pelleting or inclusion of rapidly digestible ingredients, such as glucose, did not increase prevalence or size of OC lesions. Image analysis of CT scans was a reliable method to quantify the number, size, and location of OC lesions.

Clinical and radiographic changes of carpi, tarsi and interphalangeal joints of beef zebu bulls on semen collection regimen

ABSTRACT Osteoarthritis and osteochondrosis are highly correlated to reproductive failure in bulls. This study aimed to evaluate the carpal, tarsal and interphalangeal lesions in beef zebu bulls on semen collection regimen. Twenty-one beef cattle bulls, in a total of forty-one animals, were split into three age-based groups: animals from two to four years old (GI), from more than four to eight years old (GII) and above eight years old (GIII). The clinical findings were conformational changes of limbs, synovial effusion, peripheral venous engorgement of joints and prolonged decubitus. The total population showed moderate clinical manifestation and radiographic score. The GIII presented more severe joint lesions. Carpi and tarsi regions had discrete to difuse osteophytosis, subchondral cysts, cartilaginous flaps, bone incongruence and fragmentation, osteitis, and ankylosis. Interphalangeal joints presented osteophytosis, distal phalanx osteitis and enthesophytosis. The digital radiographic examination allowed full identification of articular lesions and their clinical correspondences, besides the positive correlation between age, body weight and radiographic score.

Early detection of health and welfare compromises through automated detection of behavioural changes in pigs

Early detection of health and welfare compromises in commercial piggeries is essential for timely intervention to enhance treatment success, reduce impact on welfare, and promote sustainable pig production. Behavioural changes that precede or accompany subclinical and clinical signs may have diagnostic value. Often referred to as sickness behaviour, this encompasses changes in feeding, drinking, and elimination behaviours, social behaviours, and locomotion and posture. Such subtle changes in behaviour are not easy to quantify and require lengthy observation input by staff, which is impractical on a commercial scale. Automated early-warning systems may provide an alternative by objectively measuring behaviour with sensors to automatically monitor and detect behavioural changes. This paper aims to: (1) review the quantifiable changes in behaviours with potential diagnostic value; (2) subsequently identify available sensors for measuring behaviours; and (3) describe the progress towards automating monitoring and detection, which may allow such behavioural changes to be captured, measured, and interpreted and thus lead to automation in commercial, housed piggeries. Multiple sensor modalities are available for automatic measurement and monitoring of behaviour, which require humans to actively identify behavioural changes. This has been demonstrated for the detection of small deviations in diurnal drinking, deviations in feeding behaviour, monitoring coughs and vocalisation, and monitoring thermal comfort, but not social behaviour. However, current progress is in the early stages of developing fully automated detection systems that do not require humans to identify behavioural changes; e.g., through automated alerts sent to mobile phones. Challenges for achieving automation are multifaceted and trade-offs are considered between health, welfare, and costs, between analysis of individuals and groups, and between generic and compromise-specific behaviours.

Automatic detection of lameness in gestating group-housed sows using positioning and acceleration measurements

Lameness is an important issue in group-housed sows. Automatic detection systems are a beneficial diagnostic tool to support management. The aim of the present study was to evaluate data of a positioning system including acceleration measurements to detect lameness in group-housed sows. Data were acquired at the Futterkamp research farm from May 2012 until April 2013. In the gestation unit, 212 group-housed sows were equipped with an ear sensor to sample position and acceleration per sow and second. Three activity indices were calculated per sow and day: path length walked by a sow during the day (Path), number of squares (25×25 cm) visited during the day (Square) and variance of the acceleration measurement during the day (Acc). In addition, data on lameness treatments of the sows and a weekly lameness score were used as reference systems. To determine the influence of a lameness event, all indices were analysed in a linear random regression model. Test day, parity class and day before treatment had a significant influence on all activity indices (P<0.05). In healthy sows, indices Path and Square increased with increasing parity, whereas variance slightly decreased. The indices Path and Square showed a decreasing trend in a 14-day period before a lameness treatment and to a smaller extent before a lameness score of 2 (severe lameness). For the index acceleration, there was no obvious difference between the lame and non-lame periods. In conclusion, positioning and acceleration measurements with ear sensors can be used to describe the activity pattern of sows. However, improvements in sampling rate and analysis techniques should be made for a practical application as an automatic lameness detection system.

Associations of conformation and locomotive characteristics in growing gilts with osteochondrosis at slaughter

Osteochondrosis (OC) and abnormalities in conformation and locomotive characteristics (CLC) have been associated with premature culling in sows. Several CLC have been suggested to be associated with OC and might help as an in vivo indicator for and increased risk of having OC. The aim of this study was to investigate associations of OC with CLC assessed at several ages in growing gilts from 2 separate experiments over the effects of dietary restriction (Exp. 1) and floor type (Exp. 2) on OC prevalence. In Exp. 1, gilts (n = 211) were subjectively assessed for CLC at, on average, 4, 9, 11, 16, and 24 wk of age. In Exp. 2, gilts (n = 212) were subjectively assessed for CLC at, on average, 4, 9, 11, 16, and 22 wk of age. Assessment was done on 10 conformation and 2 locomotive characteristics using a 9-point grading scale by 2 observers. At, on average, 27 wk of age in Exp. 1 and 24 wk of age in Exp. 2, gilts were slaughtered and the knee, elbow, and hock joints were macroscopically assessed for OC. The CLC most frequently associated with OC were O shape or X shape of the hind legs, straight or bowed hind legs, and straight or sickled hock. X-shaped hind legs were associated with OC at slaughter in the knee joint at 4, 9, and 24 wk of age and at the animal level (all joints taken together) at 4, 9, and 16 wk of age. Straight or bowed hind legs were associated with OC at slaughter in the knee joint at 4 and 11 wk of age; in the hock joint at 11 wk of age; and at the animal level at 4, 9, 11, and 22 wk of age. Straight or sickled hock was associated with OC at slaughter in the knee joint at 4 wk of age, in the hock joint at 9 and 22 wk of age, and at the animal level at 9 and 22 wk of age. Results show that several CLC assessed at several ages were associated with OC, but consistent associations of a type of CLC in every assessment could not be found. The associations of CLC with OC are, therefore, difficult to be used as an in vivo indicator of increased risk for OC.

Osteochondrosis, but not lameness, is more frequent among free-range pigs than confined herd-mates

BACKGROUND

Organic pig production is expanding and amongst the objectives of organic farming are enhancing animal health and welfare. However, some studies have reported a higher prevalence of lameness and joint condemnation at slaughter in free-range/organic pigs than in conventionally raised pigs. Organic slaughter pigs have free-range housing in which indoor and outdoor access is compulsory, while in conventional farming the pigs are commonly confined to indoor pens. The present study evaluated the effects of free-range and confined housing on lameness prevalence in a herd of 106 finisher pigs, and whether osteochondrosis and Erysipelothrix rhusiopathiae associated arthritis influences these effects. We also evaluated the association between clinical lameness during the rearing period and joint condemnations at slaughter.

RESULTS

Seventy free-range and 36 confined housed fattener pigs were scored for their gait twice during the rearing period and 848 joints were evaluated post mortem. Osteochondrosis was more frequent among free-range than confined pigs (P < 0.05), and when present it was also more severe (P < 0.001). Pigs with more numerous and more severe osteochondral lesions had their gait affected more than did pigs with fewer such lesions (P < 0.05). Hence it was a paradox that we did not detect more lameness among the free-range pigs than the confined pigs. E. rhusiopathiae associated arthritis was not diagnosed. The association between gait remarks/clinical lameness and joint condemnations at slaughter was not significant.

CONCLUSIONS

The results indicate that free-range housing may have both positive and negative effects on locomotory traits. Free-range pigs may be less clinically affected by osteochondrosis than are confined pigs. One explanation for this effect may be strengthening of joint supportive tissue and pain relief promoted by exercise. Visual gait scoring missed serious joint lesions that probably were harmful to the pigs, and should therefore not be used as a sole indicator of joint/leg health in welfare inspection of pigs. The association between gait scores and joint condemnation appeared to be poor. This study was limited to one herd, and so more and larger studies on the effects of free-range housing on lameness severity and osteochondrosis development in pigs are recommended.

Pre-clinical and clinical walking kinematics in female breeding pigs with lameness: A nested case-control cohort study

Gait profiles were investigated in a cohort of female pigs experiencing a lameness period prevalence of 29% over 17 months. Gait alterations before and during visually diagnosed lameness were evaluated to identify the best quantitative clinical lameness indicators and early predictors for lameness. Pre-breeding gilts (n= 84) were recruited to the study over a period of 6 months, underwent motion capture every 5 weeks and, depending on their age at entry to the study, were followed for up to three successive gestations. Animals were subject to motion capture in each parity at 8 weeks of gestation and on the day of weaning (28 days postpartum). During kinematic motion capture, the pigs walked on the same concrete walkway and an array of infra-red cameras was used to collect three dimensional coordinate data of reflective skin markers attached to the head, trunk and limb anatomical landmarks. Of 24 pigs diagnosed with lameness, 19 had preclinical gait records, whilst 18 had a motion capture while lame. Depending on availability, data from one or two preclinical motion capture 1-11 months prior to lameness and on the day of lameness were analysed. Lameness was best detected and evaluated using relative spatiotemporal gait parameters, especially vertical head displacement and asymmetric stride phase timing. Irregularity in the step-to-stride length ratio was elevated (deviation  ≥ 0.03) in young pigs which presented lameness in later life (odds ratio 7.2-10.8).