Toon Leroy

A computer vision method for on-line behavioral quantification of individually caged poultry

In addition to production, physiology, and health, behavior is an important issue with respect to animal welfare when evaluating novel housing systems. Behavioral characteristics are usually evaluated by audio-visual observation done by a human observer present on the scene. This method is time consuming, expensive, subjective, and prone to human error. Automated objective surveillance, by means of inexpensive cameras and image-processing techniques, has the ability to generate data that provide an objective measure of behavior, without disturbing the animals. The specific purpose of this study was to develop a fully automatic on-line image-processing technique to quantify the behavior of a single laying hen as opposed to the current human visual observation. The image-processing system is based on the principle that the classification of behavior can be translated into classification of time series of different postures of the hen. The hen’s postures can be recognized in the camera image. The classification of the hen’s behavior is performed by dynamic analysis of a set of measurable parameters, which are calculated from the images using image-processing techniques. The parameters were chosen based on their computational demands and analysis of their discriminative power regarding the different types of a specific behavior. A first implementation of the system allowed us to identify three different types of individual behavior (standing, walking, and scratching). The objective of further investigation will be the classification of up to 15 different types of behavior, such as pecking, eating, drinking, wing stretching, etc.

Evaluation of Step Overlap as an Automatic Measure in Dairy Cow Locomotion

The aim of this study was to explore the possibility of capturing cow locomotion activity by computer vision techniques and to calculate the correlation between step overlap and manually measured locomotion scores. In two experiments, a total of 208 video recordings of 85 individual lactating cows were gait scored visually by an observer. The side-view videos were recorded when cows were freely passing the experimental setup. After image processing, the imprint location, step overlap, body size, and relative step overlap were calculated. The values of automatically measured step overlap showed a high correlation with the manually measured step overlap (R2 = 0.739, p < 0.001; R2 = 0.809, p < 0.001). The maximal step overlap allowed differentiation between gait scores 1 and 3 (p = 0.032) and between gait scores 2 and 3 (p = 0.039). The difference between gait scores 1 and 2 was not significant (p = 0.079). There was a large variation between individual cows, in both the progress of lameness and the influence on step overlap. Changes in step overlap were also seen that were not matched by changes in gait score. Step overlap is a variable that shows a relationship with manual gait scores, but it is not strong enough to be used as a single classifier for lameness in all cows.

Bi-Modular Flow Characterization in Tissue Engineering Scaffolds Using Computational Fluid Dynamics and Particle Imaging Velocimetry

As part of a tissue engineering (TE) therapy, cell-seeded scaffolds can be cultured in perfusion bioreactors in which the flow-mediated wall shear stress and the nutrient transport are factors that influence in vitro proliferation and osteogenic differentiation of the seeded progenitor cells. In this study both computational fluid dynamics simulations on idealized boundary conditions and circumstances and microparticle image velocimetry measurements on realistic conditions were carried out to quantify the fluid dynamic microenvironment inside a bone TE construct. The results showed that differences between actual and designed geometry and time-dependent character of the fluid flow caused a 19% difference in average fluid velocity and a 27% difference in wall shear stress between simulations and measurements. The computational fluid dynamics simulation enabled higher resolution and three-dimensional fluid flow quantification that could be quantitatively compared with a microparticle image velocimetry measurement. The coupling of numerical and experimental analysis provides a reliable and high-resolution bi-modular tool for quantifying the fluid dynamics that represent the basis to determine the relation between the hydrodynamic environment and cell growth and differentiation within TE scaffolds.

Early signs of neurolipidosis-related behavioural alterations in a murine model of metachromatic leukodystrophy.

Arylsulfatase A (ASA)-deficient mice represent an animal model for the lysosomal storage disorder metachromatic leukodystrophy (MLD). Although the model has been applied in pathophysiological and therapeutic studies, the behavioural phenotype of ASA(-/-) mice is only partially characterized, and the most decisive outcome measures for therapy evaluation only emerge beyond 1 year of age. Presently, ASA(-/-) mice and ASA(+/-) control mice were studied at 6 and 12 months of age on an extensive battery including tests of neuromotor ability, exploratory behaviour, and learning and memory. Overt signs of ataxia were not observed in 6-month-old ASA(-/-) mice, but quantitative gait analysis during open-field exploration revealed that ASA(-/-) mice displayed increased hind base width and increased stride lengths for all paws. Their covert motor incoordination was evident in a correlation analysis which unveiled decreased harmonisation of concurrent gait parameters. For example, while ASA(+/-) controls demonstrated substantial convergence of front and hind base width (r=0.54), these variables actually diverged in ASA(-/-) mice (r=-0.37). Furthermore, various behavioural observations indicated emotional alterations in ASA(-/-) mice. Six-month-old ASA(-/-) mice also showed decreased response rates in scheduled operant responding. The present findings could provide relevant behavioural outcome measures for further use of this murine MLD model in preclinical studies.

Automated gait analysis in the open-field test for laboratory mice

In this article, an automated and accurate mouse observation method, based on a conventional test for motor function evaluation, is outlined. The proposed measurement technique was integrated in a regular open-field test, where the trajectory and locomotion of a free-moving mouse were measured simultaneously. The system setup consisted of a transparent cage and a camera placed below it with its lens pointing upward, allowing for images to be captured from underneath the cage while the mouse was walking on the transparent cage floor. Thus, additional information was obtained about the position of the limbs of the mice for gait reconstruction. In a first step, the camera was calibrated as soon as it was fixed in place. A linear calibration factor, relating distances in image coordinates to real-world dimensions, was determined. In a second step, the mouse was located and its body contour segmented from the image by subtracting a previously taken “background” image of the empty cage from the camera image. In a third step, the movement of the mouse was analyzed and its speed estimated from its location in the past few images. If the speed was above a 1-sec threshold, the mouse was recognized to be running, and the image was further processed for footprint recognition. In a fourth step, color filtering was applied within the recovered mouse region to measure the position of the mouse’s paws, which were visible in the image as small pink spots. Paws that were detected at the same location in a number of subsequent images were kept as footprints—that is, paws in contact with the cage floor. The footprints were classified by their position relative to the mouse’s outline as corresponding to the front left or right paw or the hind left or right paw. Finally, eight parameters were calculated from the footprint pattern to describe the locomotion of the mouse: right/left overlap, front/hind base, right/left front limb stride, and right/left hind limb stride. As an application, the system was tested using normal mice and mice displaying pentobarbital-induced ataxia. The footprint parameters measured using the proposed system showed differences of 10% to 20% between normal and ataxic mice.

CLIMATE CONTROL BASED ON TEMPERATURE MEASUREMENT IN THE ANIMAL-OCCUPIED ZONE OF A PIG ROOM WITH GROUND CHANNEL VENTILATION

It is known that there can be a significant temperature difference between the position of the climate controller sensor (room temperature) and the animal-occupied zone (AOZ) in a pig room. This study explores the advantages of using AOZ temperature in climate control. The objectives were: (1) to evaluate a current climate control system in a practical room with ground channel ventilation for weaned piglets by comparing AOZ and room temperature, and (2) to determine advantages of control of the heating system based on AOZ temperature by a model-based predictive (MBP) controller. Comparison of AOZ and room temperature showed that during the first 10 days of the two experimental batches, AOZ temperature was lower and showed greater fluctuations than room temperature, most likely due to the switching of the heating system (on/off). Animals close to the sensor could disturb the AOZ measurement. This was not the case during colder nights, when animals moved away from the sensor and the measured AOZ temperature was a good indicator of the air temperature around the animals. The data for those periods were suitable for use in this climate control study, but when applying the system in practice the disturbing effect needs to be prevented by better protection of the AOZ sensor. For the second objective, the course of the AOZ temperature was modeled based on data for five nights when the heating switched on and off several times (goodness of fit Rt 2 = 0.77). One of the models was integrated in a simulated MBP controller that uses the model to predict future AOZ temperature; the controller switches the heating system on before the AOZ gets too cold and off before it gets too warm. The simulated AOZ temperature was more stable during an 11 h cold period; the standard deviation was reduced from 0.44°C to 0.18°C.

Automatic analysis of altered gait in arylsulphatase A-deficient mice in the open field

In current research with laboratory animals, observing their dynamic behavior or locomotion is a laborintensive task. Automatic continuous monitoring can provide quantitative data on each animal’s condition and coordination ability. The objective of the present work is to develop an automated mouse observation system integrated with a conventional open-field test for motor function evaluation. Data were acquired from 86 mice having a targeted disruption of the arylsulphatase A (ASA) gene and having lowered coordinated locomotion abilities as a symptom. The mice used were 36 heterozygotes (12 females) and 50 knockout mice (30 females) at the age of 6 months. The mice were placed one at a time into the test setup, which consisted of a Plexiglas cage (53× 34.5 × 26 cm) and two fluorescent bulbs for proper illumination. The transparent cage allowed images to be captured from underneath the cage, so image information could be obtained about the dynamic variation of the positions of the limbs of the mice for gait reconstruction. Every mouse was recorded for 10 min. Background subtraction and color filtering were used to measure and calculate image features, which are variables that contain crucial information, such as the mouse’s position, orientation, body outline, and poßible locations for the mouse’s paws. A set of heuristic rules was used to prune implausible paw features and label the remaining ones as front/hind and left/right. After we had pruned the implausible paw features, the paw features that were consistent over subsequent images were matched to footprints. Finally, from the measured footprint sequence, eight parameters were calculated in order to quantify the gait of the mouse. This automatic observation technique can be integrated with a regular open-field test, where the trajectory and motor function of a free-moving mouse are measured simultaneously.

Using light to control activity in broiler chickens

4th weeks. The cumulative mortality due to ARVF and PH-RVF in C, T1, T2 and T3 were 3, 5, 3 and 13 respectively. The incidence of ascites on d 43 in T1, T2 and T3 were 5, 5 and 4 respectively. Different heart measurements obtained on d 43 are given in Table. The RVH, as measured by RV:TV ratio in the birds died of ARVF and PH-RVF was severe while normal ratio to severe RVH was evident in d 43 birds. A significant treatment effect was present between the cumulative mortality and incidence of ascites at d 43 (P < 0.05). As shown in the Table there was a significant (P < 0.001) treatment effect on HW, RV:TV (43 d), HW:CW and RV:CW with the increase Naþ in the diet (P < 0.001). As in previous studies, this study too reveals that the Naþ level of 0.16% in feed is safe for broiler chickens. Although NaCl supplementation up to 1.11% (0.45% Naþ) in basal diets to broilers, with ascites and mortality as the criteria, was considered safe ( Julian et al., 1992), this study shows that Naþ levels of 0.37% in the diet results a ARVF incidence of 16.6%. Also, the significant difference in cumulative ascites and the incidence of ascites at d 43; and RV:TV ratio provides evidence that present-day broilers reared under tropical conditions may be more susceptible to ARVF induced through high Naþ in the feed. The authors greatly acknowledge Professor R.J. Julian for guidance and University of Peradeniya for funding.

Automatic Detection of Lameness in Dairy Cattle – Analyzing Image Parameters Related to Lameness

The objective of this research was to develop and analyze image parameters correlated with expert gait scores that are applicable for lameness detection. Experiments were done on ILVO farm in Ghent, Belgium, in August/September 2007. A camera recorded postures and movements of lactating Holstein cows. The parameters trackway overlap, hoof step time and spine arch were calculated for 10 cows. Within the evaluation it was proven that the parameters contain information that can be related to lameness. First, results showed that each image parameter had a relation to gait scores given by experts. It shows that trackway overlap, hoof step time and spine arch are useful for lameness detection. A further goal is the development of an automatic on-line lameness detection tool after analyzing more lame cows and more image parameters.

2004 SPRING MEETING OF THE WPSA UK BRANCH PAPERS

4th weeks. The cumulative mortality due to ARVF and PH-RVF in C, T1, T2 and T3 were 3, 5, 3 and 13 respectively. The incidence of ascites on d 43 in T1, T2 and T3 were 5, 5 and 4 respectively. Different heart measurements obtained on d 43 are given in Table. The RVH, as measured by RV:TV ratio in the birds died of ARVF and PH-RVF was severe while normal ratio to severe RVH was evident in d 43 birds. A significant treatment effect was present between the cumulative mortality and incidence of ascites at d 43 (P < 0.05). As shown in the Table there was a significant (P < 0.001) treatment effect on HW, RV:TV (43 d), HW:CW and RV:CW with the increase Naþ in the diet (P < 0.001). As in previous studies, this study too reveals that the Naþ level of 0.16% in feed is safe for broiler chickens. Although NaCl supplementation up to 1.11% (0.45% Naþ) in basal diets to broilers, with ascites and mortality as the criteria, was considered safe ( Julian et al., 1992), this study shows that Naþ levels of 0.37% in the diet results a ARVF incidence of 16.6%. Also, the significant difference in cumulative ascites and the incidence of ascites at d 43; and RV:TV ratio provides evidence that present-day broilers reared under tropical conditions may be more susceptible to ARVF induced through high Naþ in the feed. The authors greatly acknowledge Professor R.J. Julian for guidance and University of Peradeniya for funding.