An Instrument to Measure Turning Gait in the Community Environment

RESNA 28th Annual Conference - Atlanta, Georgia

Glenn K. Klute, Ph.D.1,2, Ava Segal, B.S.1,2, Michael S. Orendurff, M.S.1

1Dept. of Veterans Affairs, Seattle, WA.
2Dept. of Mechanical Engineering, University of Washington, Seattle, WA.


Everyday ambulation by lower limb amputees involves turning and avoiding obstacles, yet few studies have attempted to explore how prosthetic interventions might improve turning ability. To measure turning (transverse plane rotation) during gait in the community environment, a rate gyroscope was used to record rotations at velocities up to +/- 300 degrees per second from 10 intact subjects. During straight trajectory walking, the mean output of the rate gyroscope was approximately zero with a standard deviation of 6.8 degrees per second. While walking a course within a hospital, the mean (+/- s.d.) from all ten subjects indicated the subjects were walking straight 78.5% (+/- 6.8%), turning left 12.3% (+/- 4.2%), and turning right 9.2% (+/- 2.7%) of the time. The results demonstrate this instrument can measure maneuvering in the community environment.


Lower Extremity Amputee, Gait, Maneuvering, Field Measurements


Most of what is known about lower extremity amputee gait has been discovered in traditional gait laboratories while amputees walk back and forth along a straight line. Little is known about how amputees actually use their prostheses in the community. Field measurements of activities involving amputees have revealed that prosthetic components can affect daily activity levels [1, 2]. However, the step counters used in these studies cannot identify the maneuvering tasks of complex gait such as turning or avoiding obstacles. Yet complex gait is ubiquitous; one cannot go to the bathroom, fix a meal or even get out of a car without maneuvering.

While maneuvering is commonplace in everyday activities, few studies have attempted to explain how prosthetic interventions might improve turning function and stability. Two small sample population studies have suggested that additional transverse plane compliance in the prosthetic pylon can alter gait kinematics, positively influencing residual limb health and patient comfort [3, 4]. Transverse plane pylon compliance may also influence stability. Miller [5] found fifty-two percent of lower limb amputees (n=435) reported falling in the past year. If amputee falls while turning are more likely to result in serious injury as suggested by research on elderly subjects [6], then prosthetic components that might reduce this risk (i.e., facilitate turning) are an important area for research and development.

The objective of this report is to describe the development and preliminary use of an instrument capable of detecting, recording, and differentiating between steps taken while walking along a straight trajectory and while performing various turning maneuvers. Our long term aim is to use this instrument to measure the efficacy of various prosthetic interventions in facilitating amputee maneuvering.


To measure transverse plane rotation during gait in the community environment, a micro-electro mechanical system rate gyroscope (ADXRS300, Analog Devices, MA) was connected to a laptop computer data acquisition system. This sensor provided proportional voltage output in response to rotations at velocities up to +/- 300 degrees per second where positive output corresponds to turning right. Data was sampled at 200 Hz for the duration of each trial.

Figure 1: Walking Course Within A Hospital Environment (Click image for larger view)

Ten subjects free from known gait pathology provided informed consent to participate in this study. All subjects held the laptop with attached sensor while they walked at their self-selected walking speed. To determine the amount of transverse plane rotation while walking along a straight trajectory, subjects first walked down a 25m hallway free of other pedestrians. Subjects, with a guide, later walked along a course (see Figure 1) within a hospital environment without pausing. All subjects completed the course within two minutes.

Data from the straight trajectory test were processed to calculate the transverse plane rotation rate for each subject while they walked at their self-selected speed. Inspection revealed that the peak rotation rate for straight walking of each subject rarely exceeded twice the standard deviation of the sampled sensor output. Data from the course test was then processed to determine the transverse plane rotation rates for each subject and the percentage of the course where the subject was walking straight, turning left, and turning right. The criterion for walking straight was set at +/- twice the straight trajectory standard deviation. Any signal of greater magnitude indicated the subject was turning.


Figure 2: Straight Walking Rate Gyroscope Output (Click image for larger view)
Graph plots the rate gyroscope output in degrees per second versus time while a subject walked in a straight line.  The graph reveals that nearly the entire signal is bounded by twice the standard deviation of the signal sampled during the test.

The mean output of the rate gyroscope during the straight walking test was approximately zero as expected, indicating the mean heading was a straight trajectory (see Figure 2). The standard deviation (s.d.) of this signal ranged from 3.4 to 10.5 degrees/sec with a mean value of 6.8 (+/- 1.9 s.d.) for all ten subjects (see Table 1).

The rate gyroscope output from the walking course revealed the rate of transverse plane rotation and could differentiate between walking straight, turning left, or turning right (see Figure 3). The data from all ten subjects revealed they performed turns at greater than 50 degrees per second 1.3% of the time, between 40 and 50 degrees per second 1.2% of the time, between 30 and 40 degrees per second 3.0% of the time, between 20 and 30 degrees per second 6.6% of the time, and less than 20 degrees per second 87.9% of the time. The mean (+/- s.d.) from all ten subjects indicated the subjects were walking straight 78.5% (+/- 6.8%), turning left 12.3% (+/- 4.2%), and turning right 9.2% (+/- 2.7%) of the time (see Table 1).

Table 1: Rate Gyroscope Output for Ten Subjects
Hallway Test
Mean (s.d.)
Straight s.d. Deg/sec
6.8 (1.9)
Hospital Environment Test
78.5 (6.8)
9.2 (2.7)
12.3 (4.2)


Figure 3: Hospital Course Rate Gyroscope Output (Click image for larger view)
Graph plots the rate gyroscope output in degrees per second versus time while a subject walked a course in a hospital environment (see description of Figure 1).

Transtibial and transfemoral amputee function is often limited by residual limb discomfort and instability. Everyday ambulation involves turning and avoiding obstacles. These maneuvers increase the torsional load at the residual limb-prosthetic socket interface and may result in increased discomfort and soft tissue injuries. To reduce the incidence of discomfort and injury, prosthesis manufacturers have developed rotators and pylons that allow transverse plane rotation, but these devices are not widely prescribed, leaving one to wonder whether their design provides significant benefit or if their benefit is just not well understood. Few investigators have explored their effect on kinematics [3, 4] and no data exist documenting their performance in the field.

To properly understand the benefits of rotational compliance, one must first establish the prevalence of maneuvering in everyday living. One can easily establish the number of steps a person takes in a day, but there is currently no instrument to measure how many times someone turns. Without such an instrument, one cannot test the hypothesis that transverse plane rotation pylons, or any other intervention, might facilitate (or inhibit) complex gait. The lack of an appropriate instrument to measure maneuvering during daily activities suggests a gap in our knowledge base regarding how amputees actually use their prostheses. This study demonstrated the ability to measure transverse plane rate of rotation and differentiate between turning and walking straight, indicating its potential for performing in-field turning measurements. Our current research involves creating a pager-sized instrument to be worn inside the pylon of a lower limb amputee.


 This research was funded by the Department of Veterans Affairs project # A2661C in collaboration with Cyma, Seattle, WA.

Corresponding author:

Glenn K. Klute, Ph.D.
Dept. of Veterans Affairs
1660 S. Columbian Way MS151
Seattle, WA 98108