5
Lameness detection trial in a commercial environment using an embedded microcomputer based force plate system

Tuesday, March 17, 2015: 2:45 PM
401 (Community Choice Credit Union Convention Center)
Brady M. McNeil , Department of Animal Science, Iowa State University, Ames, IA
Joseph D. Stock , Department of Animal Science, Iowa State University, Ames, IA
Julia A. Calderón Díaz , Department of Animal Science, Iowa State University, Ames, IA
Thomas D. Parsons , Department of Clinical Studies, School of Veterinary Medicine, University of Pennsylvania, Kenneth Square, PA
Anna K. Johnson , Department of Animal Science, Iowa State University, Ames, IA
Locke A. Karriker , Iowa State University College of Veterinary Medicine, Ames, IA
Suzanne T. Millman , Iowa State University College of Veterinary Medicine, Ames, IA
Steven J. Hoff , Agricultural and Biosystems Engineering, Iowa State University, Ames, IA
Kenneth J. Stalder , Department of Animal Science, Iowa State University, Ames, IA
Abstract Text:

Visual locomotion scoring systems are commonly used to evaluate lameness; however, their reliability depends on observer training and experience. Therefore, developing a more accurate lameness detection method is needed. The embedded microcomputer based force plate system (force plate) is an objective measuring tool that uses the sow’s weight distribution to detect lameness, which has been used in lab based lameness trials with induced lameness and proven successful. The objective of this study was to identify whether the force plate differed from visual scoring in lameness detection in a commercial setting. The force plate system was installed under one of two electronic sow feeders in a dynamic group pen with 120 multiparous sows for 21 days. Force distribution was recorded continuously during the 21 days. Force recordings were made once per second beginning when feed was provided to the sow and all four feet were on their specific quadrants. The average force distribution measurement per week was calculated and a cutoff of 10% difference in force distribution between limbs was used to classify a sow as lame or non-lame for each week. Additionally, sows were visually evaluated using a 0 to 3 lameness assessment scale (0 = normal and 3= severely lame) every week by an experienced observer and then classified as non-lame (score ≤ 1) or lame (score ≥ 2). Data were analyzed using generalized estimated equations in SAS PROC GENMOD. Model included lameness scoring method, measuring week and their interaction. There was no difference in lameness classification between the two scoring methods (odds ratio = 1.0; Cl = 0.43 to 2.32; P > 0.05). There was an increase in the number of sows classified as lame, by both the force plate and visual assessment, on week 1 (Odds ratio = 1.28; Cl = 0.80 to 2.04; P < 0.05) compared to week 2, but there was no difference in the number of lame sows between weeks 2 and 3 (P > 0.05). The results of this research suggest that the force plate system can work in a commercial system to identify lameness in group housed sows. Further work is necessary to investigate how lameness progresses over time and to create a decision tree to improve force plate lameness detection accuracy when lameness is evaluated on a daily basis. 

Keywords: Sow, Lameness Detection, Weight Distribution