Page 175 - Adams and Stashak's Lameness in Horses, 7th Edition

Examination for Lameness 141

assistance is used to track, record, segment (into strides), The Q® with Lameness Locator ®
and extract relevant (to lameness) features from the The Q® (hardware) and Lameness Locator® (soft­
VetBooks.ir trajectories, most likely the asymmetry of trajectory ware) was specifically designed as an aid to the practic­
trajectories of the marked body parts. Analysis of the
ing equine veterinarian for detection and evaluation of
motion between right and left parts of the stride, can
then be used to detect and measure lameness. Although lameness in horses. Equine veterinarians and engineers
there are many different commercially available cam­ at the University of Missouri in collaboration with the
era‐based systems, they all depend on unobstructed Hiroshima Institute of Technology in Japan developed
line‐of‐site light transmission. The camera‐based kine­ the technology.
matic technique is more sensitive than subjective evalu­ The Q consists of three inertial sensors, a tablet PC for
ation for detecting asymmetry of motion because the data collection, analysis, and archiving, a sensor battery
sampling rate of the camera can exceed the temporal charger, and accessories for attaching the sensors to the
resolution of the unaided human eye by several orders horse’s body (Figure 2.130). The inertial sensors are
of magnitude. In order to collect multiple contiguous attached to the head, right forelimb pastern, and dorsum
strides at maximum spatial resolution, kinematic evalu­ of the pelvis between the tuber sacrale. Each sensor is
ation of lameness in horses is usually performed on the 1.5 inches by 1.25 inches by 0.75 inches and weighs
equine treadmill with the horse remaining within the 28 g. Vertical accelerations of the head and pelvis and
cameras’ field of view. Movement of the horse on a angular velocity of the right distal forelimb are meas­
treadmill is different than overground with longer ured and wirelessly transmitted (Bluetooth Class 1) in
stance times, earlier placement of the forelimb, increased real time to a handheld tablet computer. Range of trans­
retraction of both fore‐ and hindlimbs, and decreased mission is up to 100 m, although larger transmission dis­
11
vertical movement of the hooves and withers. Some tances can be attained with additional Bluetooth repeater
lameness that is visible (or measurable) overground technology or with more sensitive or directional anten­
may not be seen (or measured) on the treadmill, and nas. Sensor data (acceleration) is algorithmically con­
vice versa. Horizontal ground reaction forces on the verted to vertical position relative to the ground, which
limbs of horses, which are different when trotting on is the data of interest to the evaluator. Additional cus­
the treadmill, with increased braking (negative horizon­ tom‐designed algorithms are used to detect and quantify
tal in the first half of stance) and decreased propulsive forelimb and hindlimb lameness when the horse is mov­
(positive horizontal in the second half of stance), may ing in a symmetrical gait, which is usually the trot.
be the cause of this discrepancy. Newer kinematic sys­ Proper trotting strides are automatically detected by the
tems with automatic calibration and tracking of mark­ software when the horse is moving. Non‐trotting strides
ers and user‐friendly software to collect and display are automatically extracted and discarded before analy­
results greatly simplify use of this technique, but the sis. An additional sensor can be attached to the lower
requirement for dedicated space and multiple cameras waist of the rider to evaluate rider activity, to determine
to collect multiple strides with adequate spatial if the rider is sitting or posting correctly, and to assess
resolution limits this practice to research centers and how rider activity affects the measurement of lameness.
technologically advanced clinics. Most of the early A new research version that tracks movement with 9
investigations searching for the most sensitive and degrees of freedom and allows determination of accel­
accurate indicators of pain‐induced lameness were car­ eration in the global reference frame is available, allow­
ried out using these systems, and it is expected that ing flexibility in placement of sensors (number and
most of the research into development of models to location) for further kinematic research. Automatic
study equine motion for many different reasons will Interpretation and Degree of Evidence (AIDE) algo­
continue with this type of technology. rithms, developed from studies of induced lameness and
Asymmetry of motion can be more readily meas­ from contemporary clinical cases, for straight‐line eval­
ured using inertial sensors attached to the horse’s uation, lunging, flexions, effect of blocking and treat­
body. 6–8,18,21,32,34–36,42,48,54,55,57,58,60,61,64–66,69 Sensor data is ment, under‐saddle evaluation, and compensatory
wirelessly transmitted to the evaluator, so spatial resolu­ lameness, simplify interpretation of kinematic values
tion does not decrease with distance of subject from the related to lameness for the practitioner. Mobile Q, with
evaluator, and line‐of‐site requirements are relaxed. collection of data on smartphone devices and analysis in
Motion data from multiple contiguous strides in an the cloud, is currently in development.
overground setting can be collected and evaluated. Lameness Locator® algorithms were developed from
Wireless transmission of body‐mounted inertial sensor previous kinematic research and clinical testing. Best
data allows clinicians to objectively evaluate lameness in sensor type and locations were determined by data min­
horses in a natural clinical environment. For these rea­ ing of accumulated motion data from groups of sound
sons, using body‐mounted inertial sensors to measure horses, horses with naturally occurring lameness, and
lameness is more intuitive and practical than other horses with various models of induced lameness. Vertical
methods. In one study inertial sensor‐derived lameness head and pelvic acceleration are converted to vertical
measures agreed more closely with blinded subjective position in space and separated into components by cus­
evaluation of experts than results from a stationary tom error‐correction algorithms. Random and nonperi­
18
force plate. In another study, body‐mounted inertial odic movement is discarded, minimizing the negative
sensors identified an induced lameness in horses trotting effect of asymmetry due to nonclinical reasons.
in a straight line before a consensus of expert evaluators Remaining periodic movement is separated into move­
performing full lameness evaluations. The remainder ment due to lameness at stride frequency (the lameness
45
of this section describes one system in use today. component) and natural vertical motion at twice stride