74 Gait Analysis.PDF

C H AP T E R 7 4 Gait Analysis Bryan T. Torres The use of gait analysis in the veterinary community has steadily increased during the past few decades. Various techniques have been used for the evaluation of patients with musculoskeletal and/or neurologic conditions as well as for the evaluation o...

C H AP T E R 7 4 Gait Analysis Bryan T. Torres The use of gait analysis in the veterinary community has steadily increased during the past few decades. Various techniques have been used for the evaluation of patients with musculoskeletal and/or neurologic conditions as well as for the evaluation of various treatments for surgical and nonsurgical conditions. Veterinarians face challenges in the application of gait analysis techniques, many of which are not encountered in the study of human gait. Veterinarians have a diverse array of species requiring care, ranging from the large equine quadruped to the smallest of winged bipedal avian species. Furthermore, within many species (e.g., the dog), there can be very diverse body morphologies. Regardless of species, the interest in and use of gait analysis are becoming more commonplace, and it is important that veterinarians are familiar with gait analysis techniques. The proper use of these techniques, as well as an understanding of how to interpret the subsequent results, is critical for the advancement of our profession and patient care. Gait analysis is typically divided into two main categories: kinetics and kinematics. Kinetic gait analysis is the study of forces generated during movement. Kinematic gait analysis is the study of motion irrespective of masses or forces. Most modern gait analysis laboratories have the ability to gather information about one or both of these areas. The equipment for kinetic and kinematic analyses is discussed later in the chapter. The presence of dedicated space for small-animal gait analysis is becoming increasingly sought after in veterinary facilities. Laboratory size varies among current veterinary gait analysis centers. Fortunately, kinetic gait analysis can be performed successfully in areas with limited space. However, kinematic gait analysis requires a larger laboratory space. Currently, there are no uniform guidelines regarding laboratory size and equipment layout to help veterinarians plan the ideal gait laboratory. It is imperative that careful planning occurs in order to maximize the usefulness of the available space. When planning a veterinary gait laboratory, one should attempt to secure as quiet and spacious a location as possible. Gait and the Gait Cycle Quadrupedal gait encompasses a series of coordinated movements of the feet and limbs.46 Gaits are typically divided into two main categories: symmetrical gaits and asymmetrical gaits.46,108 The most commonly studied gaits are the symmetrical gaits, which include the trot, walk, and pace. What makes these gaits “symmetrical” is that the limb movement on one side of the body is repeated on the opposite side. For example, during the trot, a diagonally located thoracic limb and pelvic limb (e.g., right thoracic limb and left pelvic limb) support the body. This is then followed by the alternative pair of diagonally located limbs. Asymmetrical gaits include the canter, transverse gallop, and rotary gallop. These gaits are “asymmetrical” because limb movement on one side is not exactly repeated on the opposite side, thus increasing the complexity of interpretation and making them uncommon in gait assessment studies.  The complete gait cycle consists of two main phases: the stance phase and the swing phase. The stance phase is when the foot is in contact with the ground, and the swing phase is when the foot is in the air. The combination of one stance phase and the subsequent swing phase equals one complete gait cycle for an individual limb. Kinetic Gait Analysis Kinetic gait analysis is the study of forces created by and during the stance phase of the gait cycle. When an animal's foot/paw contacts the ground, it exerts a force on the ground and the ground reacts with an equal and opposite force (Newton's third law), the ground reaction force. A ground reaction force represents the resultant force of all local forces acting on the foot/paw of the animal. However, it should be remembered that ground reaction forces represent the force from an individual limb and are not specific for any one joint.46 Every time a foot/paw makes contact with a measurement device, such as a force plate, a three-dimensional force vector is produced and can be completely described by nine distinct quantities: three orthogonal components of the force vector (Fx, Fy , Fz ) that describe the direction of the force, three spatial components (x, y, z) describing the location of the force vector on the plate, and three orthogonal moments (Mx, My , Mz ).30 The direction of force is the most widely reported and studied. Although the ground reaction force actually has the multiple components just described, the term “ground reaction force” is often used to refer to only the directional force vectors: (1) vertical (Fz ), (2) craniocaudal (Fy ), and (3) mediolateral (Fx) (Figure 74.1). The vertical force (Fz ) is the most commonly evaluated force, followed by craniocaudal force (Fy ). The mediolateral force (Fx) is not often utilized for comparative purposes due to the small amount of force exerted in that direction as well as inconsistent results.27,120  FIGURE 74.1 Depiction of ground reaction forces. Vertical ground reaction forces (Fz) are represented on the Z axis, craniocaudal (Fy) on the Y axis, and mediolateral forces (Fx) on the X axis (dotted line). (From Gordon-Evans, WJ: Gait analysis. In Tobias KM, Johnston SA, editors: Veterinary surgery small animal, St Louis, 2012, Elsevier/Saunders.) Force and weight are closely related, and an understanding of this relationship is important for veterinarians interested in kinetic gait assessment. Clinicians routinely refer to and record the “weight” of patients in kilograms. However, a kilogram is a measure of mass. Weight is a measure of an object's mass in relation to a gravitational force (e.g., Earth's gravity), and its units are the newton (N). The equation used to derive weight [weight (W) = (mass of a patient in kilograms) × (the acceleration of gravity on Earth in m/s2)] is a more specific form of the general force equation [force (F) = (mass of an object in kilograms) × (the acceleration of the object in m/s2)].30 Therefore, weight is a force and on Earth is equivalent to the effect of Earth's gravity (g = 9.81 m/s2) on the mass of an object (e.g., dog). Because the gravitational force (g) is part of this calculation, there can be slight changes in an object's weight depending on latitude and altitude. Today, most modern electronic scales used in homes, hospitals, and veterinary facilities report “mass” in kilograms because they undergo strict factory calibration and account for the acceleration of Earth's gravity in their calculations. However, as just discussed, weight is a force, which means it is a vector quantity with a magnitude and a direction. This is important in understanding kinetic gait evaluation. While standing still on the ground, weight (gravity's effect on mass) is directed downward, toward the Earth's center. However, as one will notice when standing on solid soil, one does not sink into the ground because there is an equal and opposite upward force exerted from the Earth (Newton's third law), a ground reaction force. In that example, a person's weight (a force) and the subsequent ground reaction force are equivalent and therefore offset each other, leading to a resultant vector force of zero. However, ground reaction forces produced during ambulation are often greater than or less than the magnitude of the animal's weight. It is the changing magnitude (values) of these “equal and opposite” ground reaction forces that is evaluated during dynamic kinetic gait evaluation. Data from force plates are collected over time and therefore are often initially displayed graphically for quick visual interpretation (Figure 74.2). The vertical force (Fz ) is the largest force, and in dogs it is represented as a simple “bell shape” during the trot and has an “M shape” during the walk. The reason for this difference in shape is the speed at which events are occurring during the trot compared to the slower walk.46 The craniocaudal force (Fy ) is the next largest force. This force helps define two periods of forward motion during stance phase: braking (deceleration) and propulsion (acceleration).27,46 Braking occurs at the beginning of the stance phase, when an animal is initially contacting the ground and a decrease in forward momentum is occurring. This is characterized by positive force values. Propulsion occurs later in stance during the time when the limb is pushing off and propelling the animal forward. This period is characterized by negative force values. In normal dogs, greater braking than propulsion occurs in the thoracic limbs, and in the pelvic limbs, there is greater propulsion than braking. The mediolateral force (Fx) is the smallest force in dogs that are walking or trotting in a straight line (see Figure 74.2). This reduced amplitude, as well as large variation in results for the mediolateral force compared to the vertical and craniocaudal forces, has resulted in little attention to this force.46 However, with the increased interest in canine agility and athletic events, in which dogs are undergoing rapid directional changes, it is possible that evaluation of the mediolateral force may prove valuable. FIGURE 74.2 Graphic representation of ground reaction forces at a trot and at a walk. The red line represents the vertical force (Fz), the blue line represents the craniocaudal force (Fy), and the yellow line represents the mediolateral force (Fx). The peak force is labeled with a red arrowhead for the vertical (Fz) force during the walk and trot. The craniocaudal peak forces (Fy) are labeled with a blue arrowhead for the walk. The vertical impulse is depicted for the thoracic limb trot graph with a dotted area under the curve. The craniocaudal impulses are labeled for the walk with a cross-hatched area under the curve. When kinetic data are collected, the most widely reported and compared values for all three orthogonal forces (Fz , Fy , and Fx) are the peak force and impulse values. Peak force is simply the “maximum” force exerted in the respective direction (i.e., the peak of the force curve) (see Figure 74.2). The impulse value is the area under the force-time curve and takes both force and contact time of the limb into consideration. For the vertical force (Fz ), the peak and impulse values produce a singular value for each limb. Both of these values are accepted as measures of function and as indirect measures of pain.a With lameness or pain, the peak vertical force and vertical impulse are reduced due to a reduction in weight bearing on the affected limb as well as a reduction in time that weight is applied to the leg. Unlike the singular values in the vertical force, the craniocaudal force (Fy ) contains a breaking and propulsive component. The peak and impulse values for each of these two components are typically evaluated separately for each limb. Braking and propulsive forces are both reduced in animals with cranial cruciate ligament insufficiency.28,64,109,120 This is also true of animals with thoracic limb lameness; however, during a walking gait, braking is more affected than propulsion.1 Craniocaudal force has also been used to evaluate gait in animals with neurologic abnormalities.82,133 Other components of the force curves that are evaluated, but to a lesser extent, are the rising and falling slope. Although these areas of the force curve depict the ground reaction forces generated during limb contact with the force plate, they are often described as representing the loading or offloading of “weight.” The rising slope of the force curve is the period from initial contact (i.e., 0 value) to the maximum force. This region of the curve represents the loading of weight on the limb, and the slope gives the loading rate. A steeper slope represents a more rapid loading of weight on the limb, whereas a lower slope represents slower loading of weight. The falling slope is the period from maximal force to when contact with the ground ceases (i.e., 0 value). This portion of the curve represents the unloading of weight from the limb and gives the offloading rate. A steeper slope represents a more rapid offloading of weight on the limb, whereas a lower slope represents slower offloading of weight. In animals with lameness, there are changes in both the rising and the falling slope. The rising slope is reduced, consistent with a more cautious initial placement of weight on the limb (reduced loading rate). The falling slope is increased, consistent with a quicker removal of weight from the limb (increased offloading rate). Kinetic Gait Analysis Equipment Force Plate Systems Force plates use transducers to measure force. Various types of transducers are in use, including strain gauge, piezoelectric sensor, Hall effect sensors, capacitance gauge, and piezoresistive sensors. Most commercially available force plates use either strain gauge or piezoelectric sensor transducers.30 Companion animal gait laboratories can be designed to house a single force plate or multiple in-line force plates. The use of multiple in-line force plates allows researcher to collect information on a greater number of footfalls during a single pass over the force plate(s). The use of two force plates compared to a single plate reduces overall collection time and trial repetition.132 This may be beneficial in animals with orthopedic disease. Force plates can be mounted in the laboratory in several ways. Recessed or pit installation is achieved by embedding the force plate(s) into the floor so that the top of the plate(s) is flush with the remaining floor. This method is easiest during new construction when planning can occur prior to installation of the floor and force plates. In established laboratories, a section of the flooring must be removed to establish the pit and allow installation of the force plate(s). This method is labor intensive and allows for less flexibility in the future if modifications to the laboratory are needed. Floor mounting or raised installation is performed by permanently affixing the force plate(s) to the existing flooring. A raised floor (platform) is then built around the force plate(s) to a level that is flush with the top of the force plate(s). This is a simpler method of mounting force plates and as such is commonly found in small-animal gait laboratories. This method allows for increased long-term flexibility of the gait laboratory space. Freestanding installation or the use of portable force plates is also available. This method of installation or use of portable force plates is uncommon in veterinary medicine. Regardless of installation method, the metal surface of the force plates can be slippery. Because of this, the area surrounding the force plate(s), as well as the surface of the force plate itself, is typically covered with material that provides increased traction. Low-pile carpet is often used because it improves traction for the patient and is easy to maintain. Other surface coverings can be used as well. The optimal surface covering to prevent dogs from slipping during gait analysis has not been identified.68 A study evaluating ground reaction force data collected using carpet versus linoleum as a surface covering found there to be no difference.68 The measurement of patient velocity and acceleration while the patient traverses the force plate(s) is done with the aid of photoelectric switches, commonly referred to as photocells. A minimum of two photocells are required to measure patient velocity: Three photocells are required to measure acceleration: Most veterinary gait laboratories with force plate systems have three to five photocells.112 A five- photocell system has been shown to be more efficient than three photocells when collecting kinetic gait data from dogs.113 In addition, the placement of each photocell 0.5 m above the gait platform has been shown to accurately measure truncal velocity in dogs.112 Each photocell emits an invisible photoelectric beam.58 They are positioned 0.5 to 3 m apart on one side of the gait platform (over the area where the force plates are located). On the other side of the gait platform, opposite each photocell is a reflector to reflect back the photoelectric beam. The time at which each beam is broken by the animal is used to calculate both average velocity and acceleration. Because gait data are altered by large variations in velocity and acceleration, parameters for velocity and acceleration are established a priori for each study and are subsequently used as part of the criteria to accept or reject individual gait trials. Temporary structural changes to the walkway may need to be considered when obtaining gait data from animals with shorter stride lengths, such as small-breed dogs and cats. Ideally, a force plate should be of a size that allows for contact of one limb at a time. Standard-sized force plates work well for medium- and large-breed dogs. However, in smaller breeds of dogs, the dimensions of a standard-sized force plate are comparatively larger. This relative increase in plate length in combination with a shorter stride length typical of smaller dogs and cats results in multiple and simultaneous paw strikes on a single force plate. However, a method of altering a force plate's surface contact area to accommodate animals with shorter stride lengths has been described.69 Alterations to the gait platform in this manner allow for the collection of data for dogs with reduced stride lengths. This method resulted in the collection of force measurements that were consistent with the unaltered force platform, except for the peak propulsive force. Further evaluation of this technique is needed, but it is possible that this method may prove valuable for allowing collection of gait data from animals with smaller body sizes. In gait laboratories with limited space, treadmills with embedded force plates have been utilized. In general, the use of a treadmill with concurrent force plate measurement provides for a rapid collection of a large quantity of data with the use of minimal laboratory space. Unfortunately, dogs have to be trained or habituated to properly use a treadmill—thus, this method may prove more challenging. The use of a treadmill with embedded force plates has been compared with use of a standard force plate system.14 That study found that both systems recorded similar peak vertical forces for sound and lame dogs at a trot. However, the treadmill with embedded force plates did not allow for the evaluation of mediolateral and craniocaudal forces, and frequent overlap of the thoracic limb and pelvic limb paw strikes occurred. A separate study was able to eliminate this overlap by measuring individual paw strikes with the use of four force plates embedded in a treadmill.12 Although vertical force (Fz ) measurements on force plate– equipped treadmills may be comparable to those on standard force plate systems, the ability to acquire consistent and accurate information regarding the other forces (Fy and Fx) is uncertain. These limitations must be taken into account when considering treadmill use for collection of kinetic data. Pressure Walkway Systems The use of pressure walkway systems has been evaluated for gait assessment in veterinary medicine. They have been used for the evaluation of dogs with cranial cruciate ligament disease61,130; in dogs with hip dysplasia129; to follow patients undergoing total hip replacement surgery86; and for gait evaluation in cats, sheep,78 and pigs.101 Pressure walkways have a strong appeal for clinical use due to their portability and ease of storage. They are typically rolled up for storage and/or transportation and can then be unrolled for use in a wide array of locations. Pressure walkways are available in various lengths, depending on the researchers need and/or use and availability from the manufacturer. All of these factors have helped drive the clinical appeal and use of these systems, especially in locations without the benefit of dedicated laboratory space for gait analysis. The size of the patients in the study population should also be considered when deciding on the use of a pressure walkway versus a force plate system. The overall structure of pressure walkways lends themselves to the easy evaluation of animals of a wide array of sizes and body morphology. With force plate systems, the evaluation of smaller animals with shorter stride lengths can prove challenging without physical alteration to the gait platform69 due to problems in isolating or separating overlapping footfalls on a single plate. This is not a problem with pressure walkways because they can record information for each individual footfall. In addition to the ability to collect multiple gait cycles in one pass, this leads to quick acquisition of individual limb data even at low to moderate speeds—a consistent problem with traditional force plate systems.87 It is important to identify differences between pressure walkways and the more traditional force plate systems. Researchers must carefully consider what measurements are needed and what measurements can be acquired from each system. One of the most substantial differences between the two systems is the acquisition of force data. Force plate systems allow for the recording of direct force in newtons (N) in three directions: vertical, crainiocaudal, and mediolateral. Pressure walkways measure and report pressure only in the vertical direction because they are unable to measure pressure in shear directions (craniocaudal and mediolateral). Some walkway systems do allow the calculation of force values (vertical only) following calibration, whereas others do not. A study comparing data obtained from force plates and a pressure walkway confirmed that vertical force values calculated from a pressure walkway and those directly obtained from force plates systems are not comparable.85 However, that study did find that the pressure walkway produced consistent and repeatable measurements required for the evaluation of patients over time. Although pressure walkway systems are limited in their ability to assess force, their strength may be in the easy and rapid collection of temporospatial parameters such as stride time, stance time, relative stance time (stance time/stride time), walking velocity, and the calculation of symmetry indices (described later). However, unlike the well-established force plate systems, research on the effect of common sources of variability recognized in kinetic data collection (discussed later) is lacking for pressure walkway systems. Furthermore, standard collection techniques and protocols need to be defined if pressure walkway systems are to be used for the comparison of patients at different hospital locations. Evaluation of Kinetic Gait Data and Methods Kinetic gait analysis provides clinicians and researchers with a noninvasive means of collecting objective data on the forces created between the limb and the ground during stance phase. The evaluation of these forces has proven useful in the determination of normal body weight per limb distribution23,27; the effect of drug therapy on orthopedic disease24,45,61,145; lameness associated with orthopedic disease9,28,92,95,106; and the effect of diet,104 exercise,9 and weight loss95 on lameness in dogs with osteoarthritis. Appropriate evaluation and analysis of kinetic gait data are paramount. Gait data, both kinetic and kinematic, are affected by many of the same sources of variability, and similar methods of data analysis are often used. Regardless of gait analysis type, kinetic or kinematic, one must consider analysis methods carefully. Singular measurements such as peak vertical force or vertical impulse are quickly obtained, easily compared, and can provide valuable information that is clinically useful. This type of data analysis ignores the remainder of the waveform, and subtle changes that occur throughout the stance phase may be overlooked.2 Nevertheless, evaluation of the complete waveform is uncommon in kinetic evaluation, most likely due to the well-established use of point values (e.g., peak vertical force, vertical impulse) and the fact that waveform analysis is more technically demanding and labor and time intensive. Numerous components of the ground reaction forces have been used for comparison between groups. The most common variables used to compare lameness between groups of dogs over time are peak vertical force and vertical impulse.b Additional variables, such as rising and falling slope, have also been utilized. However, although individual variables are commonly used for comparative measures, it has been suggested that using a combination of variables such as peak vertical force and rising slope52 or peak vertical force and falling slope51 may improve accuracy. Normalization of kinetic data to body weight is standard practice. This process will reduce mass- related variability of kinetic data within a study population but not completely eliminate it.149 In homogeneous populations of dogs, normalization to body weight alone may be sufficient to address variability in force-dependent ground reaction force data, such as peak force. However, further efforts may be required for heterogeneous populations in which significant size differences are present. Historically, normalization directed at addressing time-dependent variables, such as stance time and velocity, has not been performed. One study advocated that normalization address not only force- dependent but also time-dependent ground reaction force variables.3 The study concluded that normalization to body weight alone did not address variability in time-dependent variables. These variables are associated with body size; therefore, normalization to withers height was performed to address differences in body size in a heterogeneous population.149 Further research by the same group determined that regardless of normalization to both body weight and withers height, ground reaction force data collected from different breeds of dog may not be comparable.151 Kinetic gait data can be collected and evaluated during what is termed a static or dynamic trial. Dynamic trials are the most common form of data collection in veterinary gait analysis. They occur during dynamic motion—dogs walking or trotting over the force plates. Static trials, on the other hand, are less often utilized and are performed by standing the dog on the force plate(s) to evaluate weight distribution. In normal standing dogs, it is commonly acknowledged that the thoracic limbs support approximately 60% of the body weight and the pelvic limbs support the remaining 40%.27 This means that each thoracic limb is responsible for carrying 30% of the body weight and each pelvic limb 20%. Studies have shown that body conformation may alter body weight distribution.105,151 Thus, with the significant intra- and interbreed variation present in the dog population, a 60/40 body weight distribution may not always occur. During dynamic gait, the force, as a percentage of body weight, may exceed three times what is observed at stance for both thoracic and pelvic limbs. A study evaluating various breeds of dogs showed that during a trot, average peak vertical force was 115% of body weight for the thoracic limbs and 72% of body weight for the pelvic limbs.151 Dynamic gait data can be obtained while the animal is walking or trotting. Much debate has occurred regarding the merits of each gait. Both are symmetrical gaits and as such are ideal for obtaining ground reaction force data. In addition, both gaits can be used to differentiate lame from non-lame dogs. However, the trotting gait has been shown to be more sensitive and accurate for the detection of lameness in dogs with low-grade or mild lameness.150 Of further benefit is that during the trot there is no overlap of footfalls—providing for easy evaluation of individual limbs. During kinetic gait data collection, it is important that the velocity and acceleration be controlled to limit variability of the gait. It has often been suggested that narrower velocity ranges are advantageous in reducing variability in gait data. However, no consensus has been reached regarding the ideal velocity for a trot or walk in dogs. This is apparent when evaluating veterinary gait publications. Hans et al.60 noted at least 10 unique trotting velocity ranges previously established in the veterinary literature. The velocity collection windows ranged from ±0.3 m/s to ±1.0 m/s, with an average of ±0.6 m/s. In that study, variance in ground reaction force data was not closely related to the magnitude of the velocity range. Thus, the narrower ranges evaluated did not minimize variability compared to the wider ranges. Wider ranges may reduce the number of rejected trials and therefore reduce overall trial number and repetition. However, careful consideration should still be used when selecting a wide velocity range during study design. In the study by Hans et al.,60 a large number of trials (average of 29) were collected and evaluated for each dog in the study. Most gait studies collect 5 valid trials from each dog for evaluation; therefore, the effect of velocity on data variability may be more profound in typical kinetic gait studies, which utilize fewer trials. Limited information on the effect of acceleration on kinetic data is available. However, it has been demonstrated that in dogs, changes to accelerations/deceleration can alter ground reaction force measurements.25 This is most apparent in the craniocaudal force values. Acceleration is typically controlled at ±0.5 m/s2 in kinetic gait studies. The use of an animal's own limb(s) to serve as an internal control for comparison is termed a “symmetry index.” The use of symmetry indices for canine gait evaluation was first introduced in the veterinary literature by Budsberg et al.23 in 1993. Since then, the interest in symmetry indices as a metric for evaluation of gait has increased. Although this metric can be applied to any gait measurement, it has been most often applied to kinetic gait data. Different methods to calculate symmetry indices have been reported.23,150 In dogs, normal locomotion is assumed to be symmetrical, and a lack of symmetry (or asymmetry) is associated with pathologic gait. However, asymmetry may exist in normal dogs during kinetic gait analysis.23 This asymmetry can be attributed to trial-to-trial variation rather than true variation between contralateral limbs.23 Because of this, normal levels of asymmetry have been suggested as