JOINT KINEMATICS DURING WHEELCHAIR PROPULSION

Sean Shimada, MS2,3, Rick Robertson, Ph.D.1,2,3, Rory Cooper, Ph.D.1,2,3, Michael Boninger, Ph.D.1,2,3 1Division PM&R, University of Pittsburgh Medical Center, Pittsburgh, PA 15261 2Dept. Rehab. Science & Technology, University of Pittsburgh, Pgh, PA 15261 3Human Engineering Research Laboratories, Highland Dr Veterans Affairs Medical Center, Pgh, PA 15206

ABSTRACT

The purpose of this study was to describe and characterize elbow and shoulder joint kinematics in three dimensions. Six experienced wheelchair users were filmed with a three camera motion analysis system. Each subject pushed a Quickie 1 wheelchair fitted with a force sensing wheel (SMARTWheel) [1], at two speeds (3 and 5 mph). The elbow angle was analyzed in the sagittal plane, while the shoulder joint was analyzed in the sagittal and frontal plane. There was a significant difference (p<0.05) in the mean start shoulder angle and total excursion in the x-y plane and mean start and end angles in the y-z plane when analyzed between the two speeds. The identification of the kinematic components that contribute to an efficient wheelchair propulsion stroke can be used to establish a model for optimizing performance.

INTRODUCTION

The investigation of wheelchair propulsion has become increasingly popular due to the growing population of wheelchair users. The act of wheelchair propulsion has been described as the bilateral, simultaneous, repetitive motion of the upper extremities with the upper extremities primarily going through flexion and extension [2]. Few studies have investigated the joint kinematics of wheelchair propulsion, but those who have, were typically in two dimensions and often visually estimated when the hand impacted and disengaged from the pushrim. Consequently, the exclusive use of kinematic data does not provide the researcher the exact time when the hand impacts or disengages from the pushrim, which may cause inaccuracies in data analysis. With the development of the SMARTWheel [1], we can determine the precise time when the hand impacts and disengages from the pushrim, allowing us to obtain a more distinct measure between the drive and recovery phase of the wheelchair propulsion stroke. The use of the SMARTWheel, in conjunction with kinematic data enables us to obtain more accurate data, specifically related to the joint kinematics. Joint angles at impact and disengagement can be calculated, which can be further used to determine joint excursions and maximum and minimum joint angles during propulsion and recovery. The purpose of this study was to evaluate the shoulder and elbow kinematics for two speeds of wheelchair propulsion.

METHODS

Four male and two female experienced wheelchair users with spinal cord injuries, volunteered and gave informed consent for the study. Each of the subjects pushed on a Quickie 1 wheelchair fitted with the SMARTWheel [1], while secured to a wheelchair dynamometer. The SMARTWheel was used to indicate when the hand applied a force to the pushrim. The subjects pushed the wheelchair at two speeds (3 and 5 mph) for three minutes. A 60 Hz, three-dimensional camera system (Peak Performance Technologies, Inc.) was utilized to collect kinematic data at during the last minute of each trial. A synchronization pulse generated by the kinetic data collection system was utilized to ensure that the video and SMARTWheel data collection were synchronous. Highly reflective spheres used for digitizing purposes were used to identify 7 anatomical landmarks, greater trochater, acromion process, lateral epicondyle of the humerus, radial and ulnar styloid process, and the 2nd and the 5th metacarpalphalangeal joint on each subject. The start angles were defined as the time when force was exerted to the pushrim, while the end angles were the time when the force returned to the baseline. The total excursion of the joints were calculated by the difference between the end and start angles. The elbow in full extension is defined as 180 degrees with the angle decreasing as the elbow flexes. The shoulder angle in the x-y plane is defined as zero degrees when the arm is in the anatomical position. The shoulder x-y angle is positive when flexion occurs at the joint and negative when the joint goes into extension. The shoulder angle in the y-z plane is defined as 90 degrees when the arm is abducted to the horizontal position. The angle decreases as the arm adducts towards the body. The maximum and minimum angles of the elbow and shoulder joint were calculated as the absolute maximum and minimum angle of the joint during the propulsion stroke. Data were analyzed for 5 complete strokes.

RESULTS

Each of the subjects start and end elbow and shoulder angles were analyzed. The shoulder joint was analyzed in two planes, sagittal (x-y plane) and frontal (y-z plane). The total excursion angles of the joints were also analyzed. The mean start elbow angle for the five subjects was 111.86(10.69 degrees for the 3 mph speed and 110.26±10.85 degrees for the 5 mph speed. The mean end elbow angles were 119.72±12.16 and 119.66±13.18 degrees for the 3 and 5 mph, respectively. The mean total elbow excursion for the five subjects were 35.18±12.10 and 32.84±10.76 degrees for the two speeds, respectively. There was no significant difference (p>0.05) found in the mean elbow start, end, and total excursion angle when analyzed between the two speeds. The mean starting shoulder angle in the x-y plane measured -49.17±6.93 degrees for the 3 mph speed and -50.15±6.93 degrees for 5 mph. There was a significant difference (p<0.05) found in the mean starting shoulder angle when analyzed between speeds. The mean end angle measured -1.81±10.71 and 1.49±9.35 degrees for the 3 and 5 mph speeds, respectively. No significant difference (p>0.05) was found in the mean end shoulder angle in the x-y plane when analyzed across speeds. The mean total excursion of the shoulder in the x-y plane was 47.49±9.85 degrees for the 3 mph speed and 51.97±10.14 degrees for 5 mph. A significant difference (p<0.05) was found in the mean shoulder excursion angle when analyzed across the two speeds. The mean starting shoulder angle in the y-z plane measured 67.82±5.83 and 72.04±9.63 degrees for the 3 and 5 mph, respectively. The mean end angle for the 3 mph speed measured 41.26±11.88 and 45.60±10.82 degrees for 5 mph. When the mean start and end angles in the y-z plane were analyzed for the two speeds, there was a significant difference (p<0.05) found. The mean total excursions in the y-z plane were 27.42±13.94 and 26.85±14.30 degrees for the 3 and 5 mph, respectively. There was no significant difference (p>0.05) found in the shoulder excursion angle in the y-z plane when analyzed between the two speeds. When all subjects were analyzed within each speed, a significant difference (p<0.05) was found between subjects for all variables, with the exception of the shoulder excursion in the x-y plane.

The mean maximum and minimum elbow and shoulder angles were also measured and analyzed. The shoulder joint was analyzed in two planes of movement, the sagittal (x-y plane) and frontal (y-z plane). The mean maximum elbow angle measured 123.24±9.81 degrees for the 3 mph speed and 123.10±10.53 degrees for the 5 mph speed. There was no significant difference (p>0.05) found between the mean maximum elbow when analyzed across the two speeds. A significant difference (p<0.05) was found between subjects when analyzed within each of the speeds. The mean minimum elbow angle measured 88.12±9.26 and 90.62±8.22 degrees for the 3 and 5 mph speeds, respectively. There was a significant difference (p<0.05) found between the mean minimum elbow angle when analyzed between the 3 and 5 mph speeds. When all subjects were analyzed within the 3 and 5 mph speed, a significant difference in the mean minimum elbow angle was found (p<0.05) between subjects. The mean maximum and minimum shoulder angles in the x-y and y-z planes coincided with the start and end angles in their respective plane.

DISCUSSION

From the kinematic data, angle-angle plots were graphed for all subjects over five entire strokes. Three different plots were graphed: shoulder x-y angle vs. elbow angle, shoulder x-y angle vs. shoulder y-z angle, and shoulder y-z angle vs. elbow angle, in order to illustrate the sequencing of the elbow and shoulder joints during wheelchair propulsion. A representative set of angle-angle plots for the six subjects are illustrated in figures 1a-c. The shoulder x-y angle vs. elbow angle plot, figure 1a., illustrates that the subjects generally began their propulsion stroke with the shoulder in extension and elbow in flexion. From the plot, it can be seen that the subjects flexed their elbow, while bringing the shoulder into flexion during the initial phases of the stroke. After this point, the shoulder continued to flex while the elbow began to extend in a proportionatal fashion. A transition was seen in the elbow joint during the propulsion phase, which first flexed during the beginning of the stroke to a point near top-dead-center, then began to extend to the end of the stroke. This is supported by Rodgers et al. (1994) findings that EMG activity was primarily detected in the biceps brachii to top-dead-center, then activity in the triceps brachii thereafter. During the beginning of the recovery phase, the subjects primarily flexed their elbow, then to a greater extent began to a extend at the shoulder. Near the end of the recovery phase, the subjects continued to extend at the shoulder, but began to flex the elbow again to prepare for the propulsion phase. The shoulder y-z angle vs. elbow angle plot, illustrated in figure 1b., indicates that the subjects generally started their propulsion stroke with the shoulder in abduction and elbow in flexion. The initial phases of the propulsion stroke revealed that the subjects adducted at the shoulder, while slightly flexing the elbow. After this phase, a transition from flexion to extension was seen in the elbow joint, while the shoulder continued to adduct. The end phase of the propulsion stroke illustrated a larger proportion of elbow extension when compared to shoulder adduction.

Figure 1a-c. Angle-angle plots for the elbow and shoulder joints for 5 entire strokes.

The initial phase during recovery had a large contribution from elbow flexion with limited shoulder abduction. The remainder of the recovery phase showed a large increase in shoulder abduction with a small contribution from elbow flexion. The shoulder x-y angle vs. shoulder y-z angle in figure 1c., showed a proportionate amount of shoulder flexion and shoulder adduction during the propulsion phase of the stroke. The large contribution of shoulder flexion during the propulsion phase was supported by Rodgers et al. (1994) findings that the anterior deltoid was active throughout this phase. The end of the propulsion stroke exhibited a greater contribution of shoulder flexion with a limited amount of shoulder adduction. The joint angles during the recovery phase of the stroke followed the same relationship as the propulsion phase, but in reverse order.

Angle-angle plots are used to help illustrate and identify inconsistent joint couplings over multiple strokes. This can be used to determine inefficiencies in the user's stroking technique, which in turn, may expose the musculoskeletal structures in the upper extremities to repeated microtraumas. Inconsistencies such as differentially loading the joint structure and abrupt changes in joint angles from stroke-to-stroke can also lead to a dehabilitating injury. The identification of appropriate and erroneous kinematic components that contribute to wheelchair propulsion can be used to establish a model for optimizing wheelchair performance and further relate these characteristics to improve efficiency and prevent injury.

REFERENCES

[1] VanSickle, DP, et al. SmartWheel: Development of a Digital Force and Moment Sensing Pushrim. Proc 18th Annual RESNA Conference, Vancouver, BC, 1995.

[2] Davis, R, et al. The competitive wheelchair stroke. NSCA Journal, 10(3), 4-10, 1988.

[3] Rodgers, MM, et al. Biomechanics of wheelchair propulsion during fatigue. Arch Phys Ed Rehab, 75(1), 85-93, 1994.

ACKNOWLEDGMENTS

Partial funding for this research was provided by the U.S. Department of Veterans Affairs Rehabilitation Research & Development Services (Project #B686-RA) through the Edward Hines Jr. D.V.A. Hospital and the Paralyzed Veterans of America.

Sean D. Shimada Human Engineering Research Laboratories Veterans Affairs Medical Center 7180 Highland Drive, 151R-1 Pittsburgh, Pennsylvania 15206 WHEELCHAIR PROPULSION