PHOTO: BART NIJS


Producers that look at their ani- mals every day may under iden- tify the number of lame animals and may only notice severely lame animals.


score 2 = identifiable gait abnormality that has little impact on overall function, score 3 = identifiable abnormality which impairs function, score 4 = severe function impairment but still capable of walking, and score 5 = complete lameness.


Need for automatic lameness detection Early and accurate lameness detection and treatment is es- sential to avoid economic losses and welfare problems. It is also possible that animals may instinctively hide lameness as a self-preservation mechanism until it becomes severe mak- ing lameness assessment during the early stages difficult when using a visual scoring method. Additionally, producers that look at their animals every day may under identify the number of lame animals and only notice severely lame ani- mals. Lameness under-representation by producers results in fewer lame animals receiving veterinary treatment and thus a greater prevalence of lameness on the farm.


Automatic lameness detection Various scientific approaches have been used to develop a fully automated and continual lameness detection system based on kinematic analysis and behavioural change detec- tion using sensors, data and image analysis. Kinematic analy- sis measures the movement geometry, without considering causal forces and calculates different gait aspects, such as stride length, stance and swing duration. Since lameness can affect injured animal behaviour, parameters such as uneven weight distribution between the legs, greater weight shifting between legs while standing, increased lying times and bouts, reduced walking speed, along with changes in feeding behaviour and activity, can be used as lameness indicators. Sensors, such as cameras and a force platform, can be used to detect behavioural changes and lameness. Data and image analysis techniques can be used to evaluate, process and ex- tract useful information from images and improve image quality while reducing undesirable effects, such as noise and blurriness by enhancing resolution. The information obtained from an image using data and image analysis might include geometrical features, like dimensions, angles, shapes, density,


16 ▶ POULTRY WORLD | No. 5, 2022


and contrast features, such as darker or lighter regions; plus material related features such as rough or smooth surfaces.


Sensors applied in studies


In 2010, Nass and colleagues assessed locomotion deficien- cies in broiler chicken by analyzing the vertical peak force on both feet while walking on the force platform. Peak force asymmetry was found for each foot, independent of age or gait score. In 2017, Aydin and Silvera and colleagues indicat- ed that a three-dimensional camera with a depth sensor can be used as a non-invasive and non-intrusive tool for auto- matic broiler lameness assessment. They used the number of lying events, latency to lie down, speed, step frequency, step length, lateral body oscillation and daily activity as indicators of lameness. In 2018, Naas and colleagues developed an al- gorithm and software to analyze broiler gait score videos. Using the developed software, lameness can be detected in broilers under commercial rearing conditions as the dis- placement velocity can be easily measured. In 2018, Van Har- tem and colleagues showed that a camera-based monitoring tool for flock behaviour analysis can potentially warn the poultry farmer of possible gait problems in a commercial farm setting.


Valuable tool Smart broiler farming increases the farmer’s ability to maintain contact with individual animals as broiler production contin- ues to intensify. With smart broiler farming a large volume of data can be collected in short period of time which can be used to improve the prediction accuracy of lameness. A digital camera-based monitoring tool is accurate, inexpensive and re- liable. However, further data capture is needed to standardize animal tracking. Force platforms have good potential to sup- port future objective research in lameness, but low specificity can occur with handling alone. However, more measurements and especially delicate software are needed to create a robust system for the farm environment. Similarly, to develop a sound lameness detection system, data from different sources will need to be combined using a multivariate approach.


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36