Lincolnshire Post-Polio Library - A Service of The Lincolnshire Post-Polio Network
[ Reception ][ Library ][ Networking ][ Directory ][ What's New? ]

Gait Analysis Techniques

Rancho Los Amigos Gait Laboratory

JoAnne K. Gronley and Jacquelin Perry

The Journal of American Physical Therapy Assn. Vol. 63, No. 12, December 1984 1831-1838

Lincolnshire Post-Polio Library copy by kind permission of Dr. Perry

Ms. Gronley is Research coordinator and research Physical Therapist, Pathokinesiology Service, Rancho Los Amigos Hospital, 7601 E Imperial Hwy, Downey, CA 90242 (USA).
Dr. Perry is Director, Pathokinesiology Service, Rancho Los Amigos Hospital, and Professor of Orthopaedic Surgery, University of Southern California, Los Angeles, CA.

In the gait laboratory at Rancho Los Amigos Hospital, the emphasis is on patient testing to identify functional problems and determine the effectiveness of treatment programs. Footswitch stride analysis, dynamic EMG, energy-cost measurements, force plate, and instrumented motion analysis are the techniques most often used. Stride data define the temporal and distance factors of gait. We use this information to classify the patient's ability to walk and measure response to treatment programs. Inappropriate muscle action in the patient disabled by an upper motor neuron lesion is identified with dynamic EMG. Intramuscular wire electrodes are used to differentiate the action of adjacent muscles. We use the information to localize the source of abnormal function so that selection of treatment procedures is more precise. Force and motion data aid in determining the functional requirement and the muscular response necessary to meet the demand. Determining the optimum mode of locomotion and developing criteria for program planning have become more realistic with the aid of energy-cost measurements. Microprocessors and personal computer systems have made compact and reliable single-concept instrumentation available for basic gait analysis in the standard clinical environment at a modest cost. The more elaborate composite systems, however, still require custom instrumentation and engineering support.

Key Words: Biomechanics, Electrodiagnosis, Gait, Laboratories.

Walking is such a complex function that analysis of a patient's performance can follow any of several directions. To find out what the subject is doing, motion analysis is the appropriate technique to use. To answer the question why a particular gait pattern occurs, dynamic EMG and vector analysis are indicated. To determine the effectiveness of the person's gait, energy cost and stride analysis are required. Each approach involves specialized instrumentation and data processing capability and adequate testing space. Although the ideal testing situation is a facility that can do everything, few clinicians have their opportunity to work in such a facility. Consequently, gait laboratories differ in their organization. The direction taken generally is determined by the clinical environment, budget and space constraints, and interests of the investigators.

Rancho Los Amigos Hospital (RLAH) is a 400-bed inpatient rehabilitation program with a correspondingly large outpatient program. The disabling factor for most of our patients is paralysis from cerebral palsy, stroke, spinal cord injury, muscular dystrophy, and myelodysplasia. The hospital also has a large amputee service and an active arthritis program. The dominant clinical questions are "why" and "how effective." Consequently, the technical emphasis of the pathokinesiology service has been on dynamic EMG, footswitch stride analysis, and energy-cost measurements. Force plate and instrumented motion analysis are used to a more limited extent.


To meet daily clinical need, we developed a system of observational gait analysis.[1] This method is taught regularly to all incoming physical therapists and orthopedic residents. Because the staff can visually identify the patient's major gait deficits, an instrumented system becomes less important. The development of an observation system was fortunate because we could not meet the staffing load needed to reduce the film data from the three cameras that are used in comprehensive motion analysis. The necessity to transfer manually the marker locations from film to computer on a frame-by-frame basis generally takes an hour for each stride studied. The number of patients seen made this an overwhelming task.

The observational technique of gait analysis involves systematically assessing the motion pattern of each segment (foot, ankle, knee, hip, pelvis, and trunk). Notations of their occurrence and timing in the gait cycle are made on a chart that lists the 32 most common errors and indicates in which gait phases they are apt to occur.[1] From this work sheet, the sequence of composite limb postures is identified. By phasically relating the events at one joint to those occurring in adjacent segments, the observer can differentiate primary gait deficits from compensatory actions. Mechanisms that obstruct standing stability, inhibit progression, or increase energy cost are identified, and therapeutic plans are formulated.

For quantitated motion analysis, we most often use electrogoniometers.[*A] We have appropriate sizes for the knee, ankle, and subtalar joints. The design is a single-axis parallelogram that accommodates a shifting joint axis and minor deviations in alignment. The electrogoniometers for the knee and ankle are designed to cross the joint anteriorly so that measurements can be made of subjects wearing orthoses (Fig. 1). Accuracy has been determined by roentgenogram comparison. At the knee, soft tissue movement under the base plates resulting from muscle contraction and skin motion introduces an 8-degree error with 90 degrees of flexion. The ankle goniometer tracks within 2 degrees. A hip goniometer is not used. Several designs have been tested but proved to be inadequate to record true hip-joint motion. For all but the thinnest patient, the soft tissues prevent a firm grasp of the pelvis when applying the device. Instead, the proximal band encircles the waist. Consequently, the resulting data represent an undifferentiated mixture of spine and hip mobility. Because gait related to hip dysfunction involves markedly different mixtures of trunk, pelvis, and hip motion, we prefer not to record a composite number. When needed, observational analysis is used to distinguish trunk, pelvis, and thigh motion.

Fig 1. photograph 200 x 461 x 256 Fig 1. Double parallelogram goniometer that provides a continuous record of knee motion as the patient walks.

To provide automated recordings of the entire limb, the advantages of a videodisk camera[*B] and a microcomputer[*C] have been combined into a single plane system. With passive markers identifying appropriate landmarks on the limb, the subject's performance is recorded on videocassette.[*D] A technique for automatic tracking of the anatomical markers was developed. The system's output can be either stick figures of the limb's motion pattern or graphs of the individual joint actions. Use of the technique is limited by the time required to process the data (one hour).


In a patient disabled by paralysis, the primary cause of the gait deficit is inappropriate muscle action. When the cause is an upper motor neuron lesion, the response to manual tests of strength or the stretch reactions commonly are not equivalent to the muscle action generated by the more complex demands of walking. Thus, to understand the patient's problem, the clinician must use instrumentation that can record muscle function as it occurs rather than infer muscle action from clinical tests. For this purpose, we use dynamic EMG. The system is designed to identify timing and relative intensity of muscle action. Designation of muscle force has been approached only on a preliminary experimental basis.[2]

Myoelectric signals are generated as a neural stimulus spreads along the muscle fibers to initiate a contraction. Response is on an all-or-none basis, that is, each stimulus totally activates every muscle fiber in the motor unit. The resulting muscle force is proportional to the number of motor fibers involved. Recording of these activating signals constitutes the EMG. Milner-Brown and associates have confirmed at the motor-unit level that the all-or-none relationship between stimulation and force also applies to the EMG recorded.[3] This finding validates extending functional interpretations beyond simple timing. It also adds value to reading the raw recordings rather than reducing the data to a simple on and off line representing the muscle activity.

Surface electrodes are the most convenient means of recording muscle function because they are painless to apply and can be used by any clinician. Unfortunately, however, they do not differentiate the signals of adjacent muscles.[4] Consequently, only gross group activity can be determined, and even this may capture signals from adjacent groups. At RLAH, the emphasis is on differentiating the action of adjacent muscles by the use of intramuscular electrodes.

Reliance on wire electrodes is based on confidence that their recordings are representative of total muscle action. Early studies by Buchthal et al, however, raised some doubts.[5] By combining the finding that the signal of a single motor unit traveled a very short distance with the assumption that similar muscle fibers were in close proximity, they estimated the coaxial needle sampled only 25 motor units. Improved equipment increased the estimate to 75 motor units. Knowing that muscles contain several hundred motor units, kinesiologists assumed that the small sampling area of wire electrodes might not record significant muscle action. More recently, Burke et al have shown that the muscle fibers of one motor unit are widely dispersed both longitudinally and transversely within the muscle.[6,7] Although the area recorded by the wire electrode is small, the motor units included are representative of the action of the entire muscle. Consistency in our clinical data supports the anatomical finding of Burke et a1.[6,7]

The electrodes we use consist ofa pair of 50-µ, nylon-coated stainless steel wires[*E] with the distal 2 mm stripped of insulation and bent to hook in the muscle for stability. Short circuiting is avoided by making the hooked ends of different length so that the noninsulated areas do not touch. Basmajian and Stecko's technique of simultaneously inserting the two wires with a single 25-gauge needle is used.[8] Accuracy of the insertion in the desired muscle is confirmed by mild electrical stimulation through the wires combined with palpation of the muscle belly or tendon.

Because relative motion between adjacent muscles or at the subcutaneous tissue interface can be considerable, drawing extra wire into the fascial planes is necessary. Carrying the joint through a full range of motion and having the muscle contract strongly before the test are the techniques we use to gain wire mobility. This technique allows vigorous activity without tethering. A free loop of wire is formed at the point of skin penetration. With these precautions, the electrodes maintain their position within the muscle for the duration of the clinical testing. Because the post-test data of our research findings are similar to those obtained on the initial tests, this method of handling the wire has apparently overcome the displacement problems identified by Jonsson and Komi.[9]

Usually the test involves studying eight muscles. The myoelectric signals are transmitted by FM-FM telemetry[*F] to the recording instrumentation. Bandwidth of the system is 40 to 1,000 Hz with a roll-off at the low end beginning at 200 Hz. A 60-Hz notch filter excludes noise generated by the AC power. The overall system gain is 1,000. This signal processing eliminates most cross-talk (signal spread among muscles) so that the action of adjacent muscles can be clearly differentiated. The resulting data are stored on 7-channel analog tape.[*G] A printout on light sensitive paper[*H] is provided for immediate visual analysis.

Timing within the gait phase is the most common determination made from the EMG record. Using normal performance as the reference, patient's muscle control patterns are identified. Deviations from normal timing are classified as premature, prolonged, or continuous and out-of-phase action (Fig. 2). The orthopedic surgeons use this information to differentiate normal from abnormal function within muscle groups for final surgical planning of tendon releases or transfers in patients with upper motor neuron lesions (stroke, cerebral palsy, head trauma, or incomplete spinal cord injury). This technique has led to far better clinical results because normal function is not sacrificed and abnormal action is more precisely defined.

Fig 2. EMG output
Fig. 2. An EMG study of a stroke patient with inadequate knee flexion in swing (30°). Causes are out of phase activity of the rectus femoris muscle (RF) and continuous activity of the vastus intermedius muscle (VI). Although timing of the vastus medialis oblique muscle (VMO) is prolonged, it is not impeding knee flexion as activity terminates at the time of contralateral floor contact (vertical dashed line). The right footswitch signal (R. FT. SW.) identifies the intervals of swing (baseline) and stance (irregular staircase). The support sequence is heel, heel and first metatarsal with a brief period of flat foot (heel plus first and fifth metatarsals).

The physical therapists use this information to gain a clearer understanding of the problems they are treating. Of particular significance to them are the preliminary tests used both to assess the patient's neurological control and the quality of the electrode insertions. Before the walking runs, EMG recordings are made during manual muscle strength (selective control) and quick stretch (spasticity) tests and resisted flexor and extensor patterns. These data provide a basis for comparing the clinical findings with the muscle activity during walking.

From the raw EMG records, changes in the relative intensity of muscle action during the various tests and within the gait stride can be appreciated. These EMG records extend understanding of the patient's control capability. For example, a muscle that responds poorly to the manual muscle test and mass patterns in the supine position may function adequately during walking; at that time, a primitive synergy was used (Fig. 3). This finding means standard resistive equipment will not be an effective strengthening technique.

Fig 3. EMG output
Fig. 3. Use of EMG to clarify the discrepancy in muscle action between the manual muscle test and walking. A. EMG during side-lying manual test of the gluteus medius muscle (G. Med); note attempted substitution of the knee flexors for inability to abduct the limb. B. Activity level of same muscles during walking: G. MED = gluteus medius; SEMI MEMB = semimembranosus; BF, LH = biceps femoris, long head; BF, SH = biceps femoris, short head; KNEE GONI == knee goniometer; and L F. S. = left footswitch.

Dynamic EMG also is used to test the therapeutic effectiveness of physical therapy procedures. Adler et al found that approximation applied to normal subjects stimulates hip and knee extensor action and inhibits the plantar flexor muscles (unpublished data, 1976).

Quantification of the EMG to provide information on the relative intensity of muscle action is reserved for tests involving normal subjects. This limitation is imposed because exact placement of the electrode within the muscle cannot be controlled. As a result, the number of motor units sampled is unknown. To circumvent the problem of an undeterminable sample size, all functional data are normalized against a quantitated test.[10] Most often, a maximum sustained effort (highest one of four seconds) is used and the data are reported as a ratio (percent maximum). If several test situations are involved, dependence on a single reference value can be avoided by using the sum of all the data obtained with that electrode as the normalizing base.

Through quantitated EMG, functional differences among muscles considered to have a common action have been identified. For example, within the hip extensor muscle group, the medial hamstring muscles limit their intense action to limb deceleration in terminal swing.[10] Persistence of biceps femoris muscle action through the loading response is consistent with the external tibial torsion that must be restrained. Activity of the gluteus maximus and adductor magnus muscles focuses on stabilizing the hip during the loading response when an added knee flexion torque by the hamstring muscles would be undesirable.

Extending the concept of relative intensity to correlations between muscle force and the dynamic EMG record, muscle quality (force per unit EMG) and mode of contraction must be considered. The maximum isometric force obtained in the optimum joint position is reduced by velocity, a concentric effort, and positional changes.[11,12] Therefore, motion measurements and the individual's basic strength are essential to making muscle-force determinations.


The temporal and distance factors of gait are categorized by our laboratory as stride characteristics. To obtain this data, we insert appropriately sized insoles with compression-closing switches covering the areas of the heel, heads of the first and fifth metatarsals, and great toe into each shoe of the subject (or taped to the sole of the bare foot). As the subject walks a 10-m distance, timing of foot contact on the floor is recorded during the middle 6 m. This data interval, bracketed by photoelectric cells, represents the subject's steady state with the variations of starting and stopping excluded. The signals are transmitted by FM-FM telemetry to the recording equipment and subsequently presented in two forms. On the analog records, stance is represented by the duration of the footswitch activity and swing is the baseline interval. The foot-support pattern is displayed as steps in a staircase (Fig. 2). Steps of equal height designate the normal sequence of heel only (H), heel and fifth metatarsal (H,5), flat foot (H,1,5), and heel-off (1,5). Abnormal modes of foot support are differentiated by half steps (fifth or first metatarsal only or heel and first metatarsal). Toe contact is identified by an oscillating signal superimposed on the step recording. Currently, the analog footswitch record is used to define the stride timing of the other data such as EMG, electrogoniometers, or force measurements (Fig. 2).

Automatic calculation of the stride variables is attained with a microcomputer-based stride analyzer.[*I] The quantitated factors are expressed both as absolute values and as a percentage of normal values. In addition, a display of the foot-support pattern is provided (Fig. 4).

Fig 4. diagram Fig. 4. Stride analyzer footswitch patterns. A. Normal sequence. Stick figures have been added to identify the pattern of floor contact. B. Type of foot-support pattern with weak calf. Note flat foot contact and lack of heel-off. H = heel, 1 = first metatarsal, 5 = fifth metatarsal.

From the footswitch signals, a clinician can calculate nine gait measurements: velocity, stride length, cadence, single stance, initial and terminal double stance, total stance, gait-cycle duration, and the foot-support pattern.

These data are used to summarize the patient's ability to walk. Gait velocity (distance covered per minute) is the basic measurement. Relative effectiveness of different therapeutic procedures and the value of orthoses and prosthetic alignment are defined by gait velocity. Single stance time identifies the weight-bearing capability of the limb. This time has proved more accurate than total stance time because the latter measurement is a mixture of double and single support times. Transfer of body weight during the double stance period is highly variable when there is limb impairment; hence, floor contact per se may not be representative of the limb's support capability. The therapists and orthopedic surgeons at RLAH follow the clinical course of their patients by using a portable microcomputer unit for stride analysis because the clinical areas are several blocks from the gait laboratory (Fig. 5).

Fig 4. photograph 250 x 307 x 250 Fig. 5. Portable stride analysis system. Patient wearing footswitches and a memory unit on the belt. A 6-m distance is designated by the two light switches and a calculator is sitting on the cart.


The effort involved in walking and in propelling a wheelchair has been of particular concern to the RLAH clinical staff. Because most of our patients are severely disabled, choosing the optimum mode of locomotion for them is a daily challenge to the staff. These decisions have been facilitated by measuring the energy expended by the patients during these activities. The criteria for program planning are more realistic when they are based on energy measurements.

In determining the most effective way of measuring energy cost, our initial efforts with a treadmill demonstrated subjects with impaired limbs either cannot walk on a treadmill at all or must use a velocity that is slower than their customary speeds. Additionally, the walking must continue a sufficient time to assure an aerobic functional state. Thus, for representative data, the energy-cost measurements must be made on a circular, stationary walkway that allows uninterrupted travel for approximately five minutes. We use an outside, circular, concrete, 60-m track (Fig. 6). Performance data are collected with ECG electrodes for heart rate, a thermal sensor for respiratory rate, and a footswitch for gait velocity. The subject's expired air is collected in a plastic Douglas bag and then, after the test, is analyzed for volume and content of oxygen and carbon dioxide. Testing consists of an initial rest period for baseline data followed by the five-minute walk. The first three minutes are used to attain a steady aerobic state (heart and respiratory rates stabilize). Performance data and expired air are collected for the next two minutes. Comparison of heart rate during the first and last 30-second intervals of the data collection period are made to assure that a steady state was maintained. Data collected in other than a steady state underestimates the subject's energy requirements.

Fig 6. Photograph 450 x 419 x 250
Fig. 6. Child with cerebral palsy undergoing an energy-cost test.

From such testing, we have learned that most patients accommodate to their disability by slowing their gait velocity to reduce the demand more than they increase their rate of energy use.[13] Only children and young adults with sound cardiopulmonary systems and adequate arm and trunk strength to substitute for disabled legs register notable increases in their rate of energy use. Dependence on crutches is particularly costly of energy. Propelling a wheelchair is equivalent to normal walking in both velocity and energy cost.[14,15]

Testing has largely focused on the various clinical groups to determine appropriate functional criteria. Individual assessments are done periodically to determine the effectiveness of a particular orthosis. Another purpose is to demonstrate to the patient, family, or finance source, the dual need for a wheelchair for long distances and an orthosis for indoor and irregular terrain travel.


A force plate[*J] with piezoelectric crystal sensors initially was obtained as a research tool. The purpose was to measure the weight-bearing capacity of the hips of patients with arthritis. Results of baseline studies, however, demonstrated the intensity of the ground reaction forces was primarily determined by the subject's gait velocity.[16] The customary, double peak pattern found in normal walking disappeared when subjects reduced their walking speed from 80 to 60m a minute. As most patients walk at the slower speed, the only remaining variable was duration of maximum load. This is more easily obtained from the footswitch measurement of single limb support. Patients with unilateral pathology do show loading-rate differences, but they vary with the individual more than with the type of disability. A major limitation to using a force plate is the difficulty in obtaining reliable data. The entire foot and only that foot must contact the plate during the recording period. Yet, deliberate loading (targeting) will give false data. Setting the subject's starting point so that the desired foot will naturally strike the plate during the walking trial may require numerous repetitions. Arthritic or severely paralyzed patients commonly lack the endurance needed. The force plate, thus, was not considered clinically useful until the Moss visible vector system was developed.[17,18] After modifying the system so that the signals started with initial floor contact rather than later in the limb-loading period, this technique was adopted at RLAH to display normal and pathological weight-bearing patterns. It has proved to be an excellent teaching tool. A vector generator transposes the vertical and horizontal (saggital or frontal) ground reaction forces into a resultant vector that is displayed on an oscilloscope (Fig. 7). This represents the line of body weight. The height of the vector is made proportional to the load intensity. A lens magnifies the oscilloscope line to the size of the subject viewed in the camera lens. Simultaneous photography of the subject and the oscilloscope is accomplished with a half-silvered mirror. By the laborious technique of manually measuring the perpendicular distance between the vector and joint center markers from each frame of film and by multiplying these moment arms by the vector magnitude, we can determine the flexion and extension (or frontal plane abduction and adduction) torques that the muscles must control. The same information can be obtained by combining instrumented motion analysis and force-plate data, but currently we lack half of that system. Also color photography that superimposes the oscilloscope representation of the body weight line (vector) on the walking subject shows the demands for muscle control more clearly (Fig. 8).

Fig 4. Diagram Fig. 7. Diagram of the visible vector system. Through a half-silvered mirror (beam splitter), the oscilloscope display (scope) of the instantaneous vector is simultaneously recorded with the photograph of the walking subject.

Clinically, the visible vector system has been used to differentiate appropriate and inappropriate muscle action. If the vector displays a flexion torque at the knee, activity of the quadriceps femoris muscle is appropriate. Conversely, the presence of quadriceps femoris action when the vector is on the extensor side of the joint shows faulty muscle activity. A second cause for this event must be sought. Such information on muscle action has assisted interpretation of prosthetic alignment, postural patterns of the arthritic patient, and surgical planning.

Fig 4. photograph 250 x 634 x 250 Fig. 8. Vector display of the demands of walking during late midstance after total knee arthroplasty. Location of the body weight line indicates a dorsiflexion torque at the ankle necessitating soleus-muscle restraint. The slight extensor torque at the knee relieves the quadriceps femoris muscle of any work.


Gait laboratories provide therapists with an opportunity to obtain objective performance data on patients they are treating. In fact, in our clinical studies, the University of Southern California physical therapy students' participation is an established practice. The information necessary to document change or examine the problem depends on the clinical question to be answered but may be as simple as recording the foot-support pattern with different types of foot-wear or identifying the stride characteristics before and after changing foot or socket alignment in the amputee.

The clinical question may relate to a population of patients or to a particular treatment program as it is applied to a group of patients. Many questions could be asked. For instance, is constant passive motion after total knee replacement (or any other pathological category) more effective at regaining the rapid knee motion needed in gait than other postoperative treatment programs? One approach to answering this question might be by combining electrogoniometry and stride analysis.

Dynamic EMG opens a new avenue for therapists to confirm the type of muscle actions that result from the various therapeutic approaches designed to improve voluntary control by patients with neurological dysfunction. This information can be used in several ways. One use is organizing investigations to assess new procedures. A second use is assessing patient performance when it is contrary to what is customarily expected. The therapist can define the problems better and more accurately select the procedures that would be most advantageous to the patient. A third use is to identify precisely the muscle activity responsible for a particular gait pattern when more than one cause is possible. Patients with stroke, head trauma, or spinal cord injury may walk with a stiff-legged gait pattern for two reasons. They may lack adequate hip musculature to initiate a flexor pattern or the desired motion may be obstructed by persistent quadriceps femoris muscle activity (rectus femoris or the vasti muscles) in swing. In the latter situation, the therapist also would have the opportunity to learn whether the obstructive force is rectus femoris muscle participation with the flexor pattern or is spasticity of the vasti muscles. Such knowledge would permit more accurate treatment planning and prediction of outcome.

In addition to the use of patient testing laboratories, the physical therapist also is well-suited to contribute to the testing program. Because the value of dynamic EMG depends on the accurate anatomical placement of electrodes and the selection of muscles appropriate to the clinical question, the physical therapist is an integral member of the investigative team. Professional training of the physical therapist in anatomy, kinesiology, and clinical characteristics of the different types of disability are the factors critical to effective, definitive testing. The only additional knowledge the therapist need gain is a three-dimensional awareness of muscular relationships and that of the major vessels and nerves.

Informative motion analysis requires accurate anatomical placement of the skin markers or electrogoniometer. Some of the landmarks are very subtle and pathology can introduce significant change. The physical therapist is best prepared to meet these challenges. Appreciation for the functional potential of patients and accommodating the testing procedure to their capability are facilitated by the knowledge that therapists have.

In addition to making major contributions to the testing process, the therapist also can be of great value in study design and data interpretation. So many types of data can be obtained and methods of processing the outcome can be used that producing meaningful information depends on including a functional perspective that a skilled therapist is able to provide.


Observational analysis is the basic technique available to all physical therapists. To profit from a systematic approach, the therapist needs training, regular practice, a pad of recording forms, and an unobstructed area within the clinic. Formal training through a short continuing education course is the most expedient way to start, but self-study also is possible. Expertise is developed by participating regularly in a consultative gait clinic where difficult problems are assessed in an organized fashion. Everyone in attendance evaluates the patient's walk and completes a gait analysis form. The findings are compared to assure all are considering the same events. Differences in interpretations are resolved, a consensus is attained, and therapeutic recommendations are formulated.

Instrumented laboratories can be developed at several levels. The determinants are training, space, funds, and engineering support. The present state of microprocessors and personal computer systems have made it possible to develop compact and reliable instrumentation. Currently, several commercial single function units are available that allow the therapists to select the types of measurements most appropriate for their clinical practice.

A system to define the patient's stride characteristics should be the basic instrument because stride identifies the patient's walking ability (eg, velocity, stride length, and single limb support). Recording the patient's foot support pattern also is very helpful. The therapist can document the patient's current ability, the progress gained by various therapeutic procedures, the effect of an orthosis, or a cardiopulmonary conditioning program. Stride measurements also define the gait variables during collection of other data types such as motion or force. If a 20-ft (6-m) hallway is available, no additional space is needed to use a system such as a stride analyzer. This type of instrumentation costs approximately $7,000.

The addition of an exoskeleton goniometer (about $3,000) makes it possible to measure knee (and hip) motion. Such a system would be of particular value in a clinic having numerous arthritic patients or others undergoing various types of knee reconstruction. As yet, none of the commercial systems are designed for the ankle.

Availability of dynamic EMG is still difficult. Theoretically, one could develop a limited system for under $10,000 that would provide a quasi-accurate printout and rely on cables for data transmission. The output is too imprecise to carry clinical significance. The therapist must consider more sophisticated equipment for improved accuracy and precise coordination with other gait measurements such as foot-switches or motion.

Additional laboratory capability requires a much larger financial investment and engineering assistance. An EMG system with full capability not only would have amplifiers and recorders that could handle signals between 40 and 1,000 Hz but would also have filtering that could distinguish signals from adjacent muscles. For complete flexibility, the facility would need to add telemetry and a multichannel tape recorder. A system like this would approach $50,000. We anticipate that the difficulties in obtaining an effective EMG system will be overcome in the near future on completion of an EMG analyzer that is currently in progress.


This review of the RLAH gait laboratory has emphasized our clinical focus on patient care. Research projects have followed two directions. Technical developments have related to developing the footswitch, energy cost, and dynamic EMG systems. Functional research has assessed normal performance to provide baselines for interpreting pathological activities.


  1. Normal and Pathological Gait Syllabus. Downey, CA, Professional Staff Association of Rancho Los Amigos Hospital Inc. 1981.
  2. Hsu AT, Perry J, Hislop HJ, et al: Analysis of quadriceps muscle force and myoelectric activity during flexed knee stance. Abstract. Phys Ther 63:765-766, 1983.
  3. Milner-Brown HS, Stein RB, Vemm R: The orderly recruitment of human motor units during voluntary isometric contractions. J Physiol (Lond) 230:359-370. 1973. [PubMed Abstract]
  4. Perry J, Easterday CS, Antonelli DJ: Surface versus intramuscular electrodes for electromyography of superficial and deep muscles. Phys Ther 61:7-15, 1981. [PubMed Abstract]
  5. Buchthal F, Guld C, Rosenfaick P: Multi-electrode study of the territory of a motor unit. Acta Physiol Scand 39:83-104, 1957.
  6. Burke RE, Tsairis P: Anatomy and innervation ratios in motor units of cat gastrocnemius. J Physiol (Lond) 234:749-765, 1973. [PubMed Abstract]
  7. Burke RE, Levine DN, Salcman M, et al: Motor units in cat soleus muscle: Physiological, histochemical and morphological characteristics. J Physiol (Lond) 238:503-514, 1974. [PubMed Abstract]
  8. Basmajian JV, Stecko GA: A new bipolar in-dwelling electrode for electromyography. J Appl Physiol 17:849, 1962.
  9. Jonsson B, Komi PV: Reproducibility problems when using wire electrodes in electromyographic kinesiology. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Basel, Switzerland, Karger, 1973, vol 1, pp 540-546.
  10. Lyons K, Perry J, Gronley JK, et al: Timing and relative intensity of hip extensor and abductor muscle action during level and stair ambulation: An EMG study. Phys Ther 63:1597-1605, 1983. [PubMed Abstract]
  11. Komi PV: Relationship between musde tension, EMG and velocity of contraction under concentric and eccentric work. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Basel, Switzerland, Karger, 1973. vol 1, pp 596-606.
  12. Smidt GL: Biomechanical analysis of knee flexion and extension. J Biomech 6:79-82, 1973. [PubMed Abstract]
  13. Waters RL. Perry J, Antonelli DJ, et al: Energy cost of walking of amputees: The influence of level of amputation. J Bone Joint Surg [Am] 58:42-46, 1976. [PubMed Abstract]
  14. Waters RL, Perry J, Antonelli DJ, et al: Energetics: Application to the study and management of locomotor disabilities. Orthop Clin North Am 9:351-377, 1978. [PubMed Abstract]
  15. Waters RL, Campbell J, Thomas L, et al: Energy cost of walking in lower-extremity plaster casts. J Bone Joint Surg [Am] 64:896-899, 1982. [PubMed Abstract]
  16. Skinner SR: The correlation between gait velocity and rate of lower-extremity loading and unloading. Bulletin of Prosthetic Research 18(1):303-304, Spring, 1981.
  17. Cook TM, Cozzens BA, Kenosian H: Real-time force line visualization: Bioengineering applications. In: Advances in Bioengineering. New York, NY, American Society of Mechanical Engineers, 1977.
  18. Cook TM, Cozzens BA: The effects of heel height and ankle-foot-orthosis configuration on weight line location: A demonstration of principles. Orthotics and Prosthetics 30(4):43-46, 1976.

*A. Pathokinesology Laboratory, Rancho Los Amigos Hospital, 7601 E Imperial Hwy, Downey, CA 90242.

*B. Rotary Shutter Camera RSC-1050 and Video Disc Recorder SVM-1010. AVIDD Electronics Co. 2210 Bellflower Blvd, Long Beach. CA 90815.

*C. Apple II. Gateway Computer Center. 11470 South St, Cerritos. CA 90701.

*D. JVC Model CR 6060U Video Cassette Recording System. JVC Industries Company, 1011 W Artesia Blvd, Compton. CA 90220.

*E. .002" Stablohm 800, California Fine Wire Co, PO Box 446, Grover City, CA 93433.

*F. Model 2600, Bio Sentry Telemetry Inc, 20720G Earl St, Torrance, CA 90503.

*G. Ampex FR 1300, Ampex Corp, 401 Broadway, Redwood City, CA 84063.

*H. Honeywell 2106 Visicorder, Honeywell, Inc, 7037 Havenhurst Ave, Van Nuys, CA 91407.

*I. Footswitch Stride Analyzer, B & L Engineering, 9618 Santa Fe Springs Rd #8, Santa Fe Springs. CA 90670.

*J. Measuring Platform, Kristal Instrument Corp, 2475 Grand Island Blvd, Grand Island, NY 14072.

[ Top of Article ]

[ Catalogue (content) ][ Catalogue (source) ]

[ Reception ][ Library ][ Networking ][ Directory ][ What's New? ]

The Lincolnshire Post-Polio Network
Registered Charity No. 1064177
An Information Service for Polio Survivors and Medical Professionals

All enquiries, book requests, medical article requests, membership fees, items for newsletters and donations to
The Secretary, Lincolnshire Post-Polio Network
PO Box 954, Lincoln, Lincolnshire, LN5 5ER United Kingdom
Telephone: +44 (0)1522 888601
Facsimile: +44 (0)870 1600840
Web Site:

The Lincolnshire Post-Polio Network takes great care in the transcription of all information that appears at this site. However, we do not accept liability for any damage resulting directly or otherwise from any errors introduced in the transcription. Neither do we accept liability for any damage resulting directly or otherwise from the information available at this site. The opinions expressed in the documents available at this site are those of the individual authors and do not necessarily constitute endorsement or approval by the Lincolnshire Post-Polio Network.

© Copyright The Lincolnshire Post-Polio Network

Copyright is retained by The Lincolnshire Post-Polio Network and/or original author(s). Permission is granted to print copies of individual articles for personal use provided they are printed in their entirety. Links from other Internet WWW sites are welcome and encouraged. We only ask that you let us know so that we can in future notify you of critical changes. Reproduction and redistribution of any articles via any media, with the exception of the aforementioned, requires permission from The Lincolnshire Post-Polio Network and where applicable, the original author(s).

Document preparation: Chris Salter, Original Think-tank, Cornwall, United Kingdom.
Document Reference: <URL:>
Created: 29th March 1999
Last modification: 30th January 2010.

Valid HTML 4.0!