Kinematic and Kinetic Evaluation of the C-Leg

Evaluation of the C-Leg 1

Kinematic and Kinetic Evaluation of the C-Leg

Microprocessor Controlled Knee Unit

Dana Craig

California State University Long Beach

Introduction and Rationale

Introduction

Transfemoral amputees have historically been shown to walk slower, use significantly more energy, as well as adopt an asymmetrical gait pattern as compared to people without impairments. (Fisher, & Gullickson, 1978; Boonstra, Schrama, Fidler, & Eisma, 1994; Skinner, & Effeney, 1985; Waters, & Yakura, 1989) It has been shown that more proximal amputations exacerbate these physiological and kinematic complications (Waters, Perry, Antonelli, & Hislop, 1976), as do issues of diabetes and dysvascularized residual limbs (Pinzur, Gold, Schwartz, & Gross, 1992). Accordingly, prosthetic leg designs for transfemoral amputees would ideally allow for optimal gait biomechanics while requiring minimal energy expenditure. Of great concern in the prosthetic fitting process is choosing a knee unit that provides the necessary stability throughout the gait cycle while not inhibiting the patient in their activities of choice. To this end, the hip, knee and ankle joints must coordinate for efficient walking mechanics. Hence, walking dynamics in transfemoral amputees have been previously studied with a focus on the effects of joint component variation and prosthetic designs. (Murray, Sepic, Gardner, & Mollinger, 1980; Michael, 1999; Romo, 2000) The introduction of another microprocessor controlled hydraulic knee unit, referred to as the Computer Leg (C-Leg), has brought up clinical questions about the validity of its mechanical contributions to gait.

Just as the clinical validity of microprocessor regulated prosthetic componentry is still unclear, definitive prescription criteria for the C-Leg has not yet been announced or published. The existing criteria for prescription of the C-Leg to veterans was issued September 2000 through a joint communiqué from the Chief Consultant, Prosthetic & Sensory Aids Strategic Health Care Group (SHCG) and the Chief Consultant, Rehabilitation SHCG (Downs & Anderson, 2000; Downs, 2000). The standard established for issuance consists of the following: 1) the ability to ambulate with a variable cadence and 2) the ability to ambulate for 400 yards. Given the substantially greater cost of the C-Leg knee unit, wide spread justification would increase the overall cost of a transfemoral prosthesis. While these are appropriate preliminary criteria there is need for clarification in order to maximize patient function while keeping prosthetic services fiscally responsible.

With all technologically oriented medical advances it is imperative to validate the effectiveness of the procedure or device. However, even though investigations into microprocessor regulated prosthetic knee units are still in their infancy these devices are being more frequently offered as an alternative to conventional prostheses to the population at large. The problem is that, there is still an insufficient amount of quantifiable data to support the usage of these devices. Of principle interest is how the C-Leg knee unit performs under both simulated "real world" and laboratory conditions as compared to a conventional hydraulic or pneumatic prosthetic knee unit. Once the benefits and limitations of these devices have been identified the field as a whole will be in a better position to define patient indications and contraindications, as well as parameters for clinical issuance.

Purpose of the Study

The specific objectives for the proposed research are to identify differences between and within prostheses on the affected and sound sides for the following quantities: a) magnitude and timing of maximum knee flexion, b) magnitude and timing of maximum hip flexion, c) vertical oscillation of pelvis, d) magnitude and timing of maximum and minimum vertical ground reaction forces, e) magnitude and timing of anterior/posterior ground reaction forces, f) braking and propulsion impulse (anterior/posterior direction), g) heart rate as an estimate of energy expenditure, and h) stance phase temporal variables (e.g. time from non-affected toe strike to non-affected heel off). Based on these objectives the following hypotheses have been developed:

1)The C-Leg will reduce vaulting as evidenced by a longer period of time spent from foot flat to heel off as compared to the conventional prosthesis.

2)The subjects will exhibit a decrease in energy expenditure while using the C-Leg for overground, incline, and decline ambulation as compared to the previous prosthesis.

3)C-Leg usage will increase the amount of knee flexion during stance phase as compared to conventional knee unit usage.

4)During stair descent there will be greater kinematic symmetry with the use of the C-Leg.

5)The time to descend a set of stairs will decrease with C-Leg usage.

Significance of Research

It is desirable to aid the transfemoral amputee in maintaining the most active lifestyle possible. In the course of doing this there are several factors that have an impact. One predominant issue is the choice of prosthetic componentry. For almost thirty years the primary option for an active transfemoral amputee was to use a knee unit that incorporated either hydraulic or pneumatic piston for swing and stance phase control. While there are variations in how this piston is implemented these knee units are still essentially constrained by their mechanical design. The incorporation of microtechnology into prosthetic knee units, however, adds another alternative. It is therefore necessary to quantify exactly how these new units differ from their predecessors both in terms of how they respond under a variety of conditions as well as how the patient perceives the device to work under those conditions.

There is also a clear need for definitive objective prescription criteria for the C-Leg, as the Veterans Administration Central Office expects the C-Leg to average approximately $40,000 each when fit by private sector prosthetists (Downs, & Anderson, 2000; Downs, 2000). Even with the achievement of cost avoidance whenever possible by using the National VA Prosthetic Gait Lab in-house limb delivery system at Long Beach the cost is still approximately $18,000 per limb. Thus, the expense continues to be significant. Componentry failures and overall limb rejection by inappropriately selected veterans is not an option. Providing specific guidelines to VA clinics for the issuance of C-Leg and establishing objective repeatable gait assessment parameters for issuance will be of enormous benefit to the veteran, Rehabilitation Medicine, and Prosthetics and Sensory Aids Service.

Literature Review

For many decades, the Mauch S-N-S Hydraulic cylinder was the only compact fluid control system for prosthetic knees that offered integrated hydraulic stance and swing phase control in one unit. (Radcliffe, & Lamoreux, 1968; Mauch, 1968) The more recent CaTech cylinders function in a similar fashion. Elimination of the stance phase flexion resistance of these devices, which is much higher than is desirable for swing phase function, occurs whenever two conditions are met simultaneously. The first condition is that the knee becomes fully extended, meaning that the piston is at the limit of its travel, and secondly, a knee hyperextension moment (or tensile force on the piston) is applied for at least one tenth of one second. (Blumentritt, Wemer, Scherer, Michael, & Schmalz,1998) During the gait cycle, this pattern typically occurs during terminal stance, but it may also sometimes happen at undesirable times, such as during the period of loading response

While the theoretical basis and commercial application of combining microprocessor control with a conventional hydraulic or pneumatic knee unit has been in existence for over twenty years, remarkably little work has been done to evaluate these devices. Currently there are two prominent commercially available devices: the Intelligent Prosthesis (IP), and the C-Leg. The C-Leg consists of a fluid controlled knee unit with hydraulic orifices, which, through means of servomotors, control the hydraulic fluid flow regulating resistance to flexion and extension at the knee. Two microprocessors monitor bending angle at the knee joint, angular velocity of the knee and bending moment (dorsiflexion torque collected by strain gages) at the tibial shank.

This data are sampled at a rate of 60 Hz, providing information to the primary microprocessor which regulates hydraulic fluid through the control orifices which allows for both swing and stance control. The IP, by comparison, is a microprocessor controlled pneumatic knee unit that provides swing phase but not stance phase control.

To date the IP has undergone the most study, with interest increasing in the novel C-Leg. The primary variable examined in all of the studies performed is the amount of energy that the amputee expends while using the microprocessor controlled knee unit versus the conventional unit. Additional preliminary study has also been given to the amount of cognitive effort required of the amputee while wearing both types of knee units. Buckley, Spence, and Solomonidisfound that two of the three subjects had significantly reduced energy expenditure while walking at various speeds on a treadmill as compared to the subjects' conventional prosthesis (Buckley, Spence, & Solomonidis, 1997). In a similar case study by Taylor, Clark, Offord, and Baxter the subject was found to have expended significantly less energy while ambulating at the fastest test speed. However, they also found that there was no significant difference in energy expenditure at the two slower speeds (Taylor, Clark, Offord, & Baxter, 1996). A longer-term study of electronically controlled knee mechanisms has reported positive results. The IP received positive subjective acceptance from the amputees studied and suggested a 10% reduction in oxygen consumption required for ambulating a given distance as compared to a conventional prosthesis (Zahedi, 1996).

In a study conducted by Chin et al. the potential for using Physiologic Cost Index (PCI) as an indicator of energy expenditure, as opposed to VO2, was investigated (Chin et al., 1999). Although not a comparative study in terms of prosthetic components this study showed that there was a significant correlation between PCI and energy consumption for each of the six walking conditions tested. Subsequently, they showed that PCI could be used as an alternate evaluation method to traditional energy expenditure measurements, which allows for a less complicated testing procedure. Another two studies both inquired into the user-aspects of the IP. In the study by Datta and Howitt, they determined that the subjects rated the IP highest in terms of walking at different speeds, long distance walking and reduction of energy consumption versus their conventional prosthesis. They also reported that all of the subjects thought of the IP as an overall “improvement or a great improvement" over their previous prosthesis (Datta, & Howitt, 1998). A study by Heller, Datta, and Howitt investigated the cognitive demands of walking with both a conventional prosthesis and the IP and found that there were no significant differences in the cognitive demands on the individuals wearing either type of prosthesis. This group also used a kinematic motion system to measure head sway during ambulation while subjects wore both the IP and a conventional limb while controlled distractions were introduced. The subjects wearing the conventional limb had a statistically significant increase in head sway compared to those wearing the IP (Heller, Datta, & Howitt, 2000).

Taken as a whole, there is a nearly complete absence of instrumented gait analysis data of the C-Leg. Such information is vital to clinicians who are seeking basic information on the performance of the C-Leg and the appropriate application of this limb. The available studies do show, at a minimum, a trend-that using a microprocessor-controlled device was able to lower energy expenditure and did have an impact on subject kinematics. However, many of these studies had very few subjects or a limited number of trials. It is also noteworthy that nearly all of these investigations only looked at energy expenditure, with little consideration for kinetic or kinematic measures. Another issue with the limited scope of testing is relevance to "real world" situations. Meaning, how would these devices perform during activities that the amputee might undertake in the course of their daily lives. This would include the use of stairs and ramps, which are significant obstacles for transfemoral amputees yet are rarely studied in the amputee population.

These points all serve to illustrate the importance of conducting this study. By objectively evaluating the C-Leg under similar conditions to what an individual with a transfemoral amputation would encounter during day-to-day activities, its true benefit could be determined. Once this has been validated it will enable the clinical community to develop a better level of standards for the provision of this device. The end result of this being that this device would be appropriately prescribed so that the individuals using it will obtain the maximum possible benefit.

Methodology

Experimental Design and Timetable

This is a combination of a quasi-experimental and repeated measures study consisting of fifteen unilateral transfemoral amputees who are ambulatory with a standard prosthesis and who are willing to try the C-Leg prosthesis. The decision to use this approach is based on several factors. One of the deciding factors in using this approach is the variability of prosthetic componentry from subject to subject, both in terms of foot unit and model of the previous knee unit. As this study is intended to represent the subject under clinically optimal conditions there will be an inherent variability in the componentry used from subject to subject and prosthesis to prosthesis. Therefore by limiting the statistical analyses to within-subject and within-condition comparisons a more clinically applicable result will be obtained. Additionally, from a calculation of sample size by the following formula it was determined that using a single subject design was a valid approach: Sample size (n) = (z*s/m)2, for a confidence interval of 95%, where z* equals 1.96, s equals one standard deviation, and m equals the mean variance. In each calculation the sample size was found to be less than one, indicating that examining each subject in and of themselves would be a valid approach. Some of the variables examined as part of this calculation included the time from non-affected foot flat to non-affected heel off, time to descend a ten step staircase, and knee flexion during mid-stance.

Subsequently the number of number of necessary trials to achieve statistical power was determined using the formulas d = (μ1-μ2)/(s(2(1-p))1/2) and D =d(N)1/2where D equals power, μ1- μ2equals the expected mean difference, s equals one standard deviation, N = number of subjects (set equal to fifteen), and p equals the correlation in the population, which was set at 0.9; an alpha of .05 was also used. It was then determined for the variables mentioned above that five and ten trials were sufficient for the outdoor and indoor/laboratory testing sessions respectively. For the variables and number of trials per condition a statistical power of .99 was calculated.

Generally, as the level of amputation becomes more proximal, the interstep variability increases (Hale, 1990), and amputees who have not adapted to their prostheses do not walk as effectively as they might with accommodation (Zahedi, Spence, Solomonidis, & Paul, 1987). To gather reliable gait data, transfemoral amputee subjects must be given adequate time to adapt to the prosthesis and become oriented to the specific test procedures and facilities. Therefore, our approach incorporates an accommodation period of at least one month for each prosthetic knee unit, as well as sufficient time for the subject to become acclimated to the testing procedures. After the accommodation period, level overground ambulation both in the laboratory setting and at an outdoor track will be performed in order to assess the knee units in terms of kinetics, kinematics, and stride characteristics. Stair descent and ascending and descending an incline will also be used to test the knee unit kinematics under specialized circumstances.

Because of these factors and the study objectives the following timeline is proposed for this study. The first three months would be used for subject recruitment and accommodation to the prosthesis if they are new C-Leg wearers (January to April 2010). Data collection, reduction, and analysis would then take approximately eight months (February to October 2010). Three months are then allocated for report and manuscript preparation and generation (October 2010 to January 2011).

Patient Selection Criteria

Subjects for this project (N=15) will be drawn from all of the V.A. amputee clinics throughout Southern California and Nevada (Veterans Integrated Service Network 22). Also, if necessary, the subject pool may be augmented by patients from throughout the nation. Potential subjects would include those patients currently utilizing a C-Leg knee unit, cleared to receive a C-Leg knee unit, or who could otherwise qualify for C-Leg issuance based on the existing criteria. Additionally, by selecting these patients we would be assured that they would be capable of ambulating at least 400 yards as well as walking at multiple cadences. An additional inclusion criteria is the successful use of a conventional hydraulic or pneumatic knee unit for at least one year previous to C-Leg fitting.