Investigation of an embedded-optical-base system's functionality in detecting signal events for gait measurements
Date
2018
Authors
Atkins, Colton A., author
Pond, Kevin R., advisor
Moorman, Valerie J., committee member
Grandin, Temple, committee member
Roman-Muniz, I. Noa, committee member
Archibeque, Shawn L., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Optical sensors have the potential to provide automated gait analysis and lameness detection in livestock. Measuring animals in motion while under field conditions is difficult for current gait analysis tools, such as plate and mat methods. This has caused a lack in commercially available systems. Additionally, a deficit of these systems and others is too much noise in their signal. Current sensor systems for static or in-motion measurements rely significantly on managing this noise as a source of error. From these problems, the primary objective of this body of work was to assess the use of an embedded-optical-base system (EOBS) and its ability to obtain real-time gait measurements from livestock. The research was composed of 3 field studies and 1 controlled study. Gait data was obtained using a commercial platform (2.4 m x 0.9 m; length x width) containing 1 EOBS. A signal-base-unit (SBU) and computer were setup near the EOBS platform by integrated cabling to collect real-time signal data. Signal fluctuation measurements (i.e., signal amplitude from hoof contact; 0 to 1 arbitrary units (au)) and kinematics (e.g., estimated speed, velocity and time duration) were recorded. The sensor detected hoof contact as signal amplitudes that could be examined in real time. Visual observations and video analyses were used for validating and classifying signal readings. The initial pilot study (field test) included 8 fistulated, crossbred steers (n = 8) tested over 1 d with 2 passes per animal over the EOBS platform. Pilot study data were used to evaluate initial signal fluctuations from animal contact. A second field study included 50 crossbred and purebred (n = 20, Angus; n = 10, Hereford; n = 20, Angus x Hereford) steers and heifers (n = 50; average BW = 292.5 kg) tested on 2 d over a 1-wk period with a total of 6 passes over the EOBS platform per animal. Steer and heifer normal walks, runs, and abnormal passes over the EOBS platform were analyzed. A third controlled study consisted of 3 mixed breed horses (n = 3) that had bilateral forelimb injections. Horses had both deep digital flexor muscles injected (1 with Botox and 1 with saline) with right and left forelimbs randomized. Horses were observed on 3 d over a 124-d period consisting of pre-treatment (baseline), post-treatment, and recovery test days with 10 passes over the EOBS platform per horse per day. Primary fluctuations, true (anomaly free) signal readings, from animal contact with the EOBS platform were analyzed. True signal readings were determined based on no influence observed from other limbs. A fourth field study consisted of 8 commercial bulls (n = 8) tested on 1 d with 3 passes over the EOBS platform per bull. Bulls were classified as either normal or abnormal in musculoskeletal structure and compared to one another to observe differences in signal fluctuation patterns. During the cattle studies, animals were not controlled and allowed to walk over the EOBS platform at their own pace. These studies formed the groundwork to determine the EOBS's functionality when animals passed over the platform. Signalment (i.e., breed, sex and age) and physiological characterizations were recorded. Temperature was also recorded for cattle field tests (e.g., min -6°C to max 4°C, respectively). For all 4 studies individual animal signal measurements were analyzed for each pass over the EOBS platform, compared to video data and classified for analysis. Results from all 4 studies showed intra- and inter-animal repeatability (qualitative observation) of observed signal readings. Though a variety of hoof contact signatures were obtained, repeating patterns were evident for both groups and individual animals. The embedded-optical-base system's (EOBS) functionality proved to be robust and operable under field trial conditions. Additionally, the signal showed extremely minimal noise. Lastly, the EOBS showed a stable baseline with clear deviations from it that could be correlated to hoof contact through video validation. Though the EOBS detected animal contact per pass, future work will investigate the system's operating readiness in accurately assessing variable gait measurements for lameness detection. Overall, data provides evidence that the embedded-optical-base system (EOBS) can detect hoof contact and differentiation between types of gait based on signal events.
Description
Rights Access
Subject
gait analysis
optics
sensors
kinematics
biomechanics
precision livestock