Acceleration Analysis To Quantify Compensation In Amputee Gait: The Effect Of Prosthetic Foot Position In A Trans-Tibial Amputee

Wilson Steeleab, Mukul Talatyb, Alberto Esquenazib
aFisher Steele Technologies
Pleasanton, California, 94526, USA
b Gait & Motion Analysis Laboratory MossRehab
Elkins Park, Pennsylvania, 19027 USA

Abstract

This study investigated how prosthetic foot angle influenced the generation of propulsion during gait. We used data from a single experienced trans-tibial amputee (TTA) and acceleration analysis (AA). The subject walked in his normal foot alignment and test conditions of 5 degrees more plantarflexed and 5 degrees more dorsiflexed. Walking speed was 1.2m/s and unchanged across test conditions. In only the dorsiflexed condition did the prosthetic side show an early stance knee extensor moment. This moment contributed to propulsion more than the sound side knee at the same time. In the baseline and plantarflexed cases, the moment from the prosthetic ankle was plantarflexor in early stance, and it contributed propulsion. In late prosthetic stance, both braking from hip and propulsion from the ankle were higher with the normal alignment than either perturbation. Continued analysis of the type described are important to understanding exactly how prosthetic alignment affects function and how individuals respond.

Introduction

Joint angles, moments and power are some of the common outcome measures used to assess walking performance. Making connections between these measures and modifications to prosthetic configuration allows associations but not cause and effect relationships to be discovered. One powerful means to relate the componentry and its configuration to these outcomes – particularly the kinematics – has been reserved for elaborate and sometimes impractical forward dynamic simulation models (Talaty 2002). Acceleration Analysis (AA) represents a means to establish a relationship between changes in joint moments to the motion it produces without having to run computer simulations (Zajac, 1989; Kepple 1997). AA essentially links joint moments to the acceleration patterns they produce through out the body. We have begun to apply this methodology to understand better how prosthetic configuration influences gait performance. Comparing changes in joint moment contributions to propulsion across trial conditions may give insight into compensation strategies.

Methods

A single experienced unilateral transtibial amputee walked in a baseline and two test conditions, where the prosthetic foot was placed in five degrees dorsiflexion or five degrees plantar flexion. The subject had time to adjust to test conditions before data were collected. At least three runs were analyzed for each condition.

Custom AA codes were used to calculate the contribution to propulsion (forward acceleration of the trunk) generated by each sagittal joint moment in the three prosthetic foot configurations. For the AA computations, a three dimensional, seven segment model consisting of 2 feet, 2 shanks, 2 thighs, and a combined Head/Arms/Trunk/Pelvis (HAT) segment was used. The ankle was modeled by a universal joint, a hinge for the knee, and a spherical joint for the hips. The foot-floor interaction was modeled as a universal joint located at the GRF whenever a force plate registered >10 N.

Translational acceleration of the HAT center of mass (COM) was our measure of propulsion; when a generated acceleration had the same polarity as COM velocity, it was termed 'propulsion'; when it was opposite, it was termed 'braking'. The internal convention was used for joint torques with ankle dorsi-flexion, knee extension, and hip flexion being positive.

Results

Figure one depicts the average of propulsion for this subject's gait in normal alignment. For the majority of stance on both sides, the ankle provides propulsion. The prosthetic side ankle torque values (avg. -1.31Nm/kg @ sound HS) average higher than the sound side (avg -1.17Nm/kg @ prosthetic HS), yet produce less average propulsion during these same points in stance. Also for the baseline condition, the prosthetic side knee lacks an extension torque at ipsilateral toe off, and the HAT acceleration for this side knee lacks the forward propulsion shown at the sound side joint. Both hips show a braking region from hip extension during initial stance, and both change to a propulsive region in late stance under hip flexion. A higher hip extensor torque (avg. -1.22Nm/kg @ sound-side toe off) on the prosthetic side vs. sound side (avg. -1.08Nm/kg @ prosthetic-side toe off) results in less forward propulsion for baseline gait.

Figure 1: Contributions to HAT horizontal translational acceleration in the baseline condition from left/right ankle/knee/hip; the prosthesis was on the right side; data are average of three trials.


Figures 2 and 3 show the average value of torque and subsequent acceleration at sound side toe off (early prosthetic stance) and sound side heel strike (late prosthetic stance) across trial conditions. This provides insight into the compensation strategy used across conditions. As the majority of differences due to compensation happen on the prosthetic side in the sagittal plane for STO and SHS, only those values are displayed.

Figure 2: How stance joint moments and their contribution to propulsion varied across test conditions during early prosthetic stance. Left column is the propulsion (braking if <0) provided by the associated joint moment in the right column.


Figure 3: How stance joint moments and their contribution to propulsion varied across test conditions during late prosthetic stance. Left column is the propulsion (braking if <0) provided by the associated joint moment in the right column.


When comparing the change in torque vs. the effective change in HAT acceleration at sound toe off, fig. 2 suggests that compensation did not occur primarily at the knee at STO. It shows that both ankle plantar flexion torque and dorsi-flexion torque can provide braking; this is possible due to changes in body orientation across conditions as a compensation strategy. The effect of the dorsiflexed position can be seen here to change the polarity of knee torque at this instant from extension to flexion. The largest hip extension torque at sound-side toe off comes in the baseline alignment, but the largest hip generated braking acceleration at the HAT comes during the dorsi-flexed trials.

At sound-side HS, a nearly equivalent ankle plantar flexor torque provides vastly different HAT propulsion (fig. 3), which infers a compensation strategy based on similar net joint torques and altered segment alignment. Prosthetic side knee torque is similar across trial conditions, and is lower than hip torque for all, but only in the dorsi-flexed case does the higher hip torque generate higher propulsion.

Discussion

A prolonged early stance dorsiflexion moment (internal convention) on the prosthetic-side increased braking in the dorsiflexed test condition. The ensuing plantarflexion moment and its contribution to propulsion were diminished. The early stance prosthetic-side knee moment showed a more normal extensor profile than in the baseline or plantarflexed conditions, which contributed to propulsion. The plantarflexion condition ankle and knee moments and function were unchanged. Prosthetic-side terminal stance ankle, knee and hip synchronize to provide propulsion while the opposite hip provides braking. Both test conditions showed increased propulsion from the terminal stance hip and increased braking from the opposite hip; the effect was stronger in the dorsiflexed condition. For the normal and baseline conditions, this model displayed an early stance braking moment due to hip extension, which has not been reported in the literature for non-amputee subjects (Kepple, 1997).

References

  1. Kepple T.M. et al. Gait & Posture 6:1-8, 1997.

  2. Zajac F.E. and Gordon M.E. Exercise in Sport and Science Reviews 17:187-230, 1989.

  3. Talaty M.C. PM&R: State of the Art Reviews 16:339-360, 2002.