Developing a high sensitivity high specificity lameness detection system for dairy cattle

Bovine lameness has emerged as the second most problematic health and welfare issue for the dairy industry. Prevalence of lameness has steadily risen from 11 to 14 percent over the past 10 years. Culling due to lameness has risen from 13 to 20 percent during the same period. Estimated annual costs of lameness are $511 million because of reproduction inefficiencies, diminished milk yield, and the increased risk of involuntary culling associated with this problem. Part of the difficulty with controlling lameness is the inability to accurately and continuously observe gaits in a herd. Visual detection, though popular, suffers from limited accuracy, reproducibility, and a high labor commitment. Producers as well as highly trained professionals underestimate the prevalence of lameness in herds by as much as 40 percent. Thus, there is a significant need in the dairy industry for automated, objective detection of lameness that will reduce lameness-related losses and generally improve the wellbeing of dairy herds. BouMatic LLC took the lead in commercialization of a one-dimensional (vertical), dual-plate, force sensing technology that can automatically detect lameness in commercial dairy herds (the StepMetrix product). This novel product detects incremental changes in vertical ground reaction forces (GRFs) that are associated with changes in clinical locomotion score and lesion pain as determined by a veterinarian. Use of the one-dimension (vertical) force sensing technology detects changes in gait attributed to lameness but it suffers from limited sensitivity in lameness prediction. Step Analysis LLC has introduced a three-dimensional (vertical, longitudinal, traverse) force sensing system and used it to demonstrate improved accuracy in gait/locomotion assessment over the one-dimensional (vertical) rodent system. Step Analysis proposes to upgrade the bovine system (StepMetrix) from one-dimensional (vertical) force sensing to three-dimensional (vertical, longitudinal and transverse) force sensing which will improve the accuracy of lameness prediction to benefit the dairy industry. Successful completion of Phase I will result in an operational three-dimensional force Sensor Module for StepMetrix and validate the hypothesis that three-dimensional sensing (vertical, longitudinal, traverse) of GRFs can distinguish lame cows from sound cows with misclassification rate of less than or equal to 15 percent. A successful Phase I will pave the way for the development of a statistical model for three-dimensional lameness detection, which will require the testing of several hundred cows. This will be the subject of a Phase II project. We anticipate that the new three-dimensional (3D) StepMetrix will have higher sensitivity and specificity values than the current single-dimensional StepMetrix. This will benefit health of the cows and improve profitability for dairy farmers resulting in the greater market acceptance of automated lameness detection system in the US and globally.