Horse is trained to trot on the treadmill
i. Presently, the horse must be trained to trot on the treadmill before we can perform a computer-assisted kinematic analysis for lameness. This training is performed in stages with the first stage involving training the horse to get onto and off the treadmill willingly and smoothly. This training is done slowly and patiently without whipping or tranquilizing. Excited or tranquilized horses do not trot in the consistent manner required for accurate kinematic analysis of lameness.
ii. The second stage of training involves finding the optimum speed at which the horse is trotting comfortably on the treadmill without goading by human handlers and while maintaining a stable position on the treadmill (i.e. not moving forwards and backwards). The speed selected is highly individual to the specific horse, being influenced strongly by age, breed, athletic ability and fitness.
iii. To date we have been able to train every horse we have tried to load onto and off and to exercise at a trot on the treadmill, despite being presented with several horses who were previously known to be difficult when loading onto horse trailers.
iv. This process of training takes a maximum of 3 days. Several horses have trained to trot comfortably on the treadmill in 1 day.
v. In the future, after full development of wireless, body-sensor system treadmill exercise will not be needed for computer-assisted kinematic evaluation of lameness.
i. The treadmill laboratory is housed within the equine clinic of the University of Missouri-Columbia (UMC) Veterinary Medical Teaching Hospital (VMTH).
ii. 3-4 high-speed (120 frames/second) cameras are used to record the horse’s movement while trotting on the treadmill.
iii. Cameras are visible-red-light-detecting and strobe activated.
iv. All cameras are situated to the right or above the horse trotting on the treadmill.
v. Cameras are connected via cable to a workstation in an adjacent control room.
i. Horses are marked with small (25 mm in diameter), soft, retroreflective spheres.
ii. These spheres are attached to small, Velcro discs glued to the horses body at the following locations.1. head at the zygomatic arch (this requires in some horses shaving a small patch of hair)
2. right forelimb hoof wall
3. dorsal pelvic midline between the tubera sacral (this requires in some horses shaving a small patch of hair)
4. right hind limb hoof wall
iii. Velcro discs on the head and pelvis are left on the horse to fall off naturally. This takes from 24 hours to 2 weeks depending on the horse’s activity level and rate of hair growth.
i. After a brief warm up of a few minutes at the previous selected speed for that horse, data is collected from the high-speed, visible-red-light-detecting cameras for 30 seconds to 2 minutes. All calculations of lameness are made from the data collected from these cameras.
ii. At the same time the horse is filmed with 2 regular-speed (30 frames/second) digital cameras from the right side and immediately from the rear. This video footage is saved as a digital file for later review if needed.
i. First the 2-dimentional position of each marker is collected using a commercially available motion analysis system, the Vicon Motion Analysis System (VMAS). Each camera provides a 2-D image of the horse’s motion for the trial.
ii. The 2-dimensional films are converted into a 3-dimentional sequence of the horse’s movement for the trial using additional VMAS software and the data is saved for further analysis.
iii. The data is then analyzed with custom-written software and a custom-developed algorithm to detect forelimb and hind limb lameness.
1. Basically the vertical movement of the head correlated with the vertical movement of the right forelimb foot is used to measure forelimb lameness and the vertical movement of the pelvis correlated with the vertical movement of the right hind limb foot is used to measure hind limb lameness.
2. A frequency-domain, signal processing technique (Fast Fourier Transformation) is used to extract stride rate information from the raw head and pelvic vertical position signals.
3. The raw signals are then deconstructed into 3 components
a. A periodic component at the frequency equal to stride rate, which is the component of the raw signal due to lameness. The amplitude of this component (A1) is proportional to the amplitude of lameness.
b. A periodic component at the frequency equal to twice the stride rate, which is the component of the raw signal due to the vertical movement of the head and pelvis due to natural body intertial changes. The amplitude of this component (A2) is specific to the horse’s natural movement and speed.
c. A non-periodic component that is random, low-frequency and unrelated to natural movement and lameness.
4. The amplitude of the lameness component for the vertical head movement (A1H) and the ratio of the amplitude of the lameness component over the amplitude of the component due to natural head vertical inertial movement (A1H/A2H) are general measures of the severity of forelimb lameness. The amplitude of the lameness component for the vertical pelvis movement (A1P) and the ratio of the amplitude of the lameness component over the amplitude of the component due to natural pelvic vertical inertial movement (A1P/A2P) are general measures of the severity of hind limb lameness.
5. Side of lameness is determined by additional custom-written software that extracts and throws away the non-periodic, random component of movement and then measures the height differences of the head and pelvis at various points in the stride using the vertical movement of the right forelimb and right hind limb to determine the start and end of stance of both forelimbs and hind limbs.
a. Differences in the minimum position of the head and pelvis during the middle of the stance phase of right and left fore and hind limbs are calculated. The right stance phase is easily seen from the collected data of the vertical position of the right forelimb and right hind limb feet. Midstance phase of the left forelimb and left hind limb is assumed to occur during the swing phase of the contralateral forelimb and hind limb.
b. Differences in the maximum position just after the end of the stance phase of the right and left limbs are calculated.
c. A positive minimum head difference indicates right forelimb lameness and a positive minimum pelvic difference indicates right hind limb lameness. In other words the head and pelvis move down less during the stance phase of the corresponding lame forelimb and hind limb.
d. Negative minimum head and pelvic differences indicate left forelimb and hind limb lameness.
e. Differences in maximum position of the head and pelvis just after the end of the stance phase of the right and left limbs are calculated.
f. A positive maximum head difference usually indicates a right forelimb lameness and a positive maximum pelvic difference indicates a right hind limb lameness. In other words the head and pelvis move up less after the stance phase of the corresponding lame forelimb and hind limb.
g. Negative maximum head and pelvic differences usually indicate left forelimb and hind limb lameness.
h. Most lameness conditions in horses that we have measured to date (natural lameness conditions and induced lameness model) cause the above perturbations in head and pelvic movement. We call lameness that cause these patterns of movement, type 1 lameness conditions and believe it is cause by lameness that is most prominent, i.e. when the pain is most acute during the stride, at the beginning (i.e. at impact) or during the middle of the stance phase of the stride (when the ground reaction force on the limb is greatest). Most lameness conditions that we have measured to date have been type 1 lameness conditions.
i. ***Occasionally, a positive head maximum difference, if not accompanied by a positive head minimum difference, is an indication of lameness in the opposite or left forelimb, and a negative head maximum difference, if not accompanied by a negative head minimum difference, is an indication of lameness in the opposite or right forelimb****. We have called this a type 2 lameness and believe that it is caused by lameness that is most prominent, i.e. when the pain is most acute during the stride, at the end of the stance phase of the stride (i.e. during breakover). We have measured a few horses with this type of lameness.
6. From the calculations of A1 and A1/A2 for the head and pelvis we obtain an overall measure for the severity of forelimb and hind limb lameness. From the calculations and signs of the mininum and maximum height differences for the head and pelvis we identify the limb involved.
7. Thresholds for A1 and A1/A2 have been established experimentally between soundness and mild forelimb and hind limb lameness.