The rapid developments in prosthetic technology in the past years have improved the potential for restoring maximum function and mobility for patients with lower-limb amputation. This is supported by the published scientific evidence on the use of microprocessor-controlled prosthetic knees.1 Therefore, the extent to which the improved functions and technical advancements of prosthetic components provide additional benefits for the user should be investigated. In addition to various clinical examination procedures, biomechanical research studies are essential to answer this question because they allow for objective evaluations. One of the general methodological problems of prosthetic biomechanical research is the relatively short time allocated for amputee subjects to adapt to the new technology before testing. Within this context, there are extensive discussions in the literature on the adaptation time after which an objective biomechanical analysis should be conducted to effectively reflect the functional characteristics of the prosthetic component to be tested. Although some studies report measuring results after a relatively short adaptation phase, for example, after 1 day,2–5 other studies assume that relatively long periods of several months are required, whereas a period of 3 months is frequently used.6–10 In most cases, the statements on the adaptation period are based on subjective assessments that were obtained from practical experience. To the authors’ knowledge, English et al.11 presented the only study in which the dependence of biomechanical parameters on the subject’s adaptation time to a functionally new prosthesis is confirmed by systematic measurements. They demonstrated that an adaptation period of 1 week is required to be able to use all functions of a new knee joint during level walking, whereas objective and valid scientific measurements should be conducted after a 3-week adaptation.11 In addition, the kinematic data presented suggested that additional time to adapt after 1 week’s use does not produce corresponding changes of the movement pattern. The study was limited by the fact that only one patient was investigated, so no statements on the general validity of the results could be made. In addition, data on the moments acting on the major joints of the prosthesis and on the sound side were missing. Because these moments are of high importance for the evaluation of the functionality of transfemoral prostheses,12–14 the scientific discussion on the required adaptation time seems to be incomplete.
This study aimed to investigate long-term effects of adaptation of lower-limb amputees to a new prosthetic technology. Within the scope of an extensive test phase, before the introduction of a new prosthetic knee joint onto the market, biomechanical tests were conducted with 10 transfemoral amputees while performing different activities of daily living after a few hours of adaptation to the device. The tests were repeated after 3 months to measure the effects of adaptation after this period of use. With this study design, we investigated our hypothesis that well-established transfemoral amputees are able to use the functions of a new prosthetic knee joint after a few hours of adaptation, and therefore, a longer adaptation period is not required to obtain reliable biomechanical test results.
For this investigation, 10 male patients with unilateral transfemoral amputation were enrolled (Table 1, demographic data). Before entering the study, the amputees had been using the C-Leg®, a microprocessor-controlled prosthetic knee joint, (Otto bock, Duderstadt, Germany), for 3 to 7 years. They were recruited from different prosthetic clinics to participate in a comprehensive biomechanical evaluation of the Genium®, a new microprocessor-controlled prosthetic knee joint (Ottobock, Duderstadt, Germany). However, in an earlier investigation, all tests of this study were conducted with the same patient group with the C-Leg. The results of this investigation were published in a previous article.15
All of the patients were aware of the possible risks, and informed consent was obtained from each subject. This study was conducted in agreement with the Georg August University of Göttingen Ethics Committee.
PROSTHETIC GENIUM KNEE JOINT
Similar to the C-Leg, a monocentric knee joint that has been investigated in numerous studies,4,6–11,13,15–18 the newly developed Genium knee joint produces the resistances required in the corresponding movement phase by an electronically controlled linear hydraulic system. Compared with the C-Leg, the novel sensor technology of the Genium offers the patient new functions.15,16,19–21
- Independent of the current joint angle, knee flexion during standing is blocked by an automatic mode that has been designed for a safe and comfortable standing.
- To approximate the natural gait pattern, the knee joint is preflexed by 4° at initial contact. The joint movement under load is damped autoadaptively. During walking on ramps, increased knee flexion is produced to provide sufficient foot clearance for swinging through of the prosthetic limb.
- A new mode enables the use of a natural step-over-step movement pattern for ascending stairs with a passive knee system.
Objective gait measurements were acquired while walking on level ground at a comfortable self-selected velocity and ascending and descending a 10° ramp and stairs.
During fitting of the Genium knee joint system, the prosthetic alignment was configured according to biomechanically orientated guidelines applicable to the C-Leg.17 The settings were maintained during the whole test phase.
Immediately after fitting, the patients were given 7 to 8 hrs to make themselves familiar with the functions and the use of the new knee joint. After this first adaptation period, the abovementioned biomechanical tests were conducted (T1). Subsequently, the patients used only the new prostheses for all activities of daily living for 3 months. After this period, all biomechanical tests were repeated (T2).
For organizational reasons, the tests of sessions T1 and T2 were conducted always in the following order: 1) level walking, 2) ascending and descending ramps, and 3) ascending and descending stairs. For ascending stairs, the subjects were explicitly asked to do this with the step-over-step pattern, if possible. To avoid exhaustion, the amputees were allowed a 20-minute rest for recovery between the analyses of the different movements.
The ground reaction forces acting during level walking were determined by two force plates (9287A; Kistler, Winterthur, CH; sampling rate of 1080 Hz). Simultaneously, the motion kinematics was recorded by monitoring passive markers with optoelectronical cameras (VICON 460; ViconPeak, Oxford, GB; sampling rate, 120 Hz). For this purpose, seven markers were attached bilaterally to the patient by an experienced biomechanical engineer according to a self-developed model18: metatarsophalangeal joint, lateral malleolus (ankle adapter screw on the prosthetic side), knee center defined by Nietert22 (knee axis on the prosthetic side), trochanter, acromion, lateral humeral epicondyle, and ulnar styloid process.
The stairwell for the test consisted of five steps, the third of which was installed on a force plate.16 To measure walking on slopes, in the middle of the ramp, a 40-cm–long element was bonded to a horizontal floor-mounted standard force plate (see above). During the foot contact with this element, the ground reaction forces were measured.17,23 Both the ramp and the stairwell were equipped with a handrail. However, the handrail was not used by the patients. The detailed test setups for walking on ramps and stairs were described in earlier studies.16,17,23
DATA PROCESSING AND DATA SELECTION
All coordinates of the marker positions were smoothed using a Woltring filter (mean squared error value, 15).24 Sagittal plane joint angles were calculated using a customized model written in Vicon BodyLanguage (version 3.5; ViconPeak, Oxford, GB) for each complete gait cycle. All angular data were calculated relative to angles recorded during standing. On the basis of kinetic and kinematic parameters, the external moments generated at the lower-limb major joints were computed using standard inverse dynamics computations25 and by means of Vicon BodyLanguage.
To present the results as clearly as possible, the focus was put on parameters that had been proven to be particularly relevant for the evaluation of the activities of daily living of transfemoral amputees in earlier studies.13–18,23 These parameters include the ground reaction force and the sagittal plane joint angles and moments at the knee and the hip joints. The vertical component of the ground reaction force of the sound leg, an indicator for compensatory mechanisms used by transfemoral amputees when walking up and down ramps and stairs,17 is discussed for these motions only.
For each investigated situation, 8 to 10 gait cycles were measured. From these single trials, mean values for each biomechanical parameter were calculated. They provided the basis for the calculation of mean group values for T1 and T2. These mean group values were used to examine tendencies of meaningful changes. To answer the specific question regarding possible changes of the interindividual variability from T1 to T2, exemplary data of all patients are discussed in a separate section.
For the quantitative comparative analysis of the test series T1 and T2, peaks of the biomechanical parameters were examined for significant differences using the Wilcoxon test. This nonparametric test was chosen because of the relatively small sampling.
The time-distance parameters did not show any significant differences between T1 and T2 (p ≥ 0.05). The mean walking velocities were 1.29 (SD, 0.12) and 1.33 (0.09) m/second for T1 and T2. The mean asymmetry of the step lengths (difference between the values on the prosthetic and the sound side, absolute values) ranged between 0.05 (0.04) (T1) and 0.06 (0.04) m (T2). In both assessments, seven patients performed the longer steps with the prosthetic side. Considering striking peaks, the quantitative analysis confirms that there are no significant differences between T1 and T2 for all parameters tested (p ≥ 0.05, Table 2).
The mean prosthetic knee angle shows that the participants were able to use the new functions of the Genium to control the stance phase.15 As an example, initial contact was performed with the Genium knee joint preflexed by approximately 4° (Figure 1A), which is not allowed with the C-Leg.15 The mean knee angle and the external moment (Figure 1C) show only minimal differences between T1 and T2.
In the biomechanical comparative analysis of T1 and T2 for the hip joint on the prosthetic side, a nearly identical mean curve of the thigh segment angle was identified (Figure 1B). The mean curves of the external moment acting on the hip joint are similar as well (Figure 1D).
For both measurements, the amputees walked down the ramp at a mean velocity of 1.11 ([0.15] T1, [0.12] T2) m/second. For T1, a mean step length asymmetry of 0.03 (0.02) m was measured. For T2, the corresponding values were 0.04 (0.03) m. This difference is not significant (p ≥ 0.05). In both assessments, in six patients, the longer step length was recorded on the prosthetic side.
The analysis of the ground reaction forces on the ramp revealed that only the deceleration peak of the anterior-posterior component Fx of the sound leg acting in early stance phase differed significantly between T1 and T2 (−10  percentage of body weight [%BW] vs. −19  %BW, p ≤ 0.01). All other differences between tested biomechanical parameters were not significant (p ≥ 0.05, Table 3). The graphical representation of the vertical component of the sound leg shows that there are no qualitative changes between T1 and T2 in the mean graphs (Figure 2E). The same applies to the mean angle of the Genium (Figure 2A). The pattern of the external moment acting on the hip joint of the prosthetic side was similar in T1 and T2 (Figure 2C).
The time-distance parameters did not show any significant differences (p ≥ 0.05). The mean walking velocities were 0.99 (0.10) m/second in T1 and 1.00 (0.10) m/second in T2. The values of the step length asymmetry were 0.05 (0.03) (T1) and 0.04 (0.03) m (T2). For this motion, the longer step length was recorded on the prosthetic side in eight patients in both assessments.
All tested parameters of the ground reaction forces were insignificant between T1 and T2 (p ≥ 0.05, Table 3). Only minimal differences were identified again in the mean graphs of the vertical component measured for the sound leg (Figure 2F). In the mean knee angle, the new function of the Genium—joint preflexion at stance phase initiation—can be clearly measured (Figure 2B). The moment acting on the hip joint of the prosthetic side showed a similar pattern in T1 and T2 (Figure 2D). The maximum value of the external flexion moment of the hip on the sound side was significantly reduced at T2 (1.44 [0.21] N m/kg [T1] and 1.24 [0.23] N m/kg [T2], p ≤ 0.05). All other tested parameters showed no significant differences (p ≥ 0.05, Table 3).
When walking down stairs, no significant differences were identified for all tested parameters (p ≥ 0.05, Table 4). The vertical component of the ground reaction force of the sound leg reflects a high degree of correspondence for T1 and T2 (Figure 3D).
The prosthetic knee joint angle with preflexed stance phase initiation confirms the unique function of the Genium (Figure 3A), in which differences between T1 and T2 were not observed (maximum flexion angle, −71° [4°] T1 and −71° [5°] T2). A similar statement applies to the thigh segment angle of the prosthetic leg; the quality of continuous flexion until the end of the step contact was very similar in both cases (Figure 3B). The maximum moments produced after step contact on the hip joint of the prosthetic side tended to increase from T1 to T2 (Figure 3C).
The analysis of all calculated peak values (Table 4) demonstrated a significant change from T1 to T2 only for the knee moment on the sound leg. The value was reduced from 0.92 (0.21) (T1) to 0.71 (0.18) N m/kg (T2, p ≤ 0.05, Table 4).
The qualitative comparison of the mean graphs of the biomechanical parameters measured revealed that substantial changes developed between T1 and T2 (Figure 4). The analysis of the kinematic values on the prosthetic side showed for T2 that the mean flexion angles of the knee and the hip joint were increased during the whole swing phase (Figures 4A, B).
This effect results from drastic changes in the measurements of six patients and was not observed with any other motion investigated (see paragraph Interindividual variability of motion patterns). Loading of the prosthetic leg that is described by the vertical component of the ground reaction force was characterized by an increased rate of loading at T2 (Figure 4E). This involved a temporarily increased mean moment at weight acceptance acting at the hip joint of the prosthetic side (Figure 4C). From approximately 20% of the gait cycle until the end of the step contact, hip joint moments both on the prosthetic and sound sides were clearly reduced at T2 compared with T1 (Figures 4C, D). A similar effect was observed for the knee joints both on the prosthetic and the sound side.
INTERINDIVIDUAL VARIABILITY OF MOTION PATTERNS
The analysis of the interindividual variability showed only small changes for level walking, walking on ramps, and descending stairs when comparing T1 and T2. However, for ascending stairs at T1, there was a very large amount of variability between individuals for most biomechanical parameters, which is clearly reduced at T2. Figure 5 shows, as an example, the knee angle for level walking (top graphs) and ascending stairs (middle graphs). Although the individual differences remain approximately constant for level walking, a large variability of the motion patterns can be seen for ascending stairs at T1, particularly in the swing phase. At T2, these differences have reduced to a great extent. These changes correlate with the changes of special parameters that give information with regard to the reproducibility of the motion patterns. As an example, the mean (SD) of the forefoot marker position of the prosthetic foot is also shown in Figure 5 (horizontal component, bottom graphs). Compared with T1, this value is significantly reduced at T2 (p ≤ 0.01).
The conducted tests include motion patterns that are of great importance and frequency in the patients’ everyday lives. Before the trial fitting with the Genium, all patients investigated had been wearing the C-Leg for several years. With the C-Leg, the motion patterns required by the patients for level walking, walking up and down ramps, and descending stairs show only minor differences compared with those required when using the Genium. For ascending stairs, however, the Genium allows the patient the completely new motion pattern of climbing stairs step over step, similar to the natural movement of nonamputees. This results in meaningful clinical and biomechanical benefits: the improved symmetry enables the amputee to ambulate more inconspicuously in public, the residual limb is relieved, and the overloads to the major joints of the sound limb that are inevitably associated with the step-to technique are reduced.16 With the C-Leg, however, an asymmetrical step-to movement pattern is unavoidable.15,16 The results for the assessments T1 (a few hours after fitting with the Genium) and T2 (after 3 months of Genium use) can be “grouped” in a systematic way similar to the motion patterns necessary for the use of C-Leg and Genium. On the one hand, the mean biomechanical characteristics of the movements that can be similarly performed with both knees reflect the technical optimizations implemented in the Genium (e.g., preflexed initial contact during level walking). On the other hand, after 3 months of Genium use, only minimal changes relative to test results found immediately after the first fitting could be detected. This observation suggests that the improved control algorithms of the Genium can be immediately and intuitively used by the patient for level walking, walking on ramps, and descending stairs.15
However, with respect to the biomechanical parameters for reciprocal stair ascent, there are clear differences in the mean curve progression between T1 and T2. This may suggest that the new motion pattern has yet to be learned. Consequently, training and thus a longer adaptation period are required to master this “new” motion in an effective and reproducible manner, taking full advantage of the technical options of the Genium. The reason for the lack of significant changes with respect to the peaks of the biomechanical parameters (Table 4), despite the different curve progressions, is found in the high interindividual variability of the motion patterns, particularly at T1 (Figure 5, mid graph). For all other movements, such a dramatic change in the interindividual variability between T1 and T2 could not be observed. An example for the approximately constant interindividual variability is the knee angle for level walking shown in Figure 5 (top graph). Within the scope of the theory of motor control,26,27 the various changes from T1 to T2 could be interpreted to suggest that the amputees need to manage several movement-dependent learning processes after having been fitted with the Genium for the first time. Consequently, the results at T2 for ascending stairs, particularly the reduced interindividual variability, give reason to assume that the necessary learning process shows a convergence to a common approach across participants. This process could indicate that the learning process has been largely completed after 3 months.
These results provide valuable information for the planning of biomechanical tests that is of high importance within the scope of comparative functional analyses of prostheses. It can be assumed that biomechanical measurements during tests of knee joints that do not show any fundamental differences between the motor patterns required to use the components to be compared (as they would appear, e.g., when changing from a brake knee to a microprocessor-controlled knee) provide reliable results as early as after a few hours of adaptation in experienced and skilled users. They should then reflect the functional characteristics of the component. In the instance of functional differences that presumably require relearning or first-time learning processes, an adaptation phase should be given. On the basis of this motion-dependent classification, only parts of the initial hypothesis could be confirmed. The data provided by this study suggest that a learning phase of 3 months is likely to be sufficient, which is in agreement with previous results and assumptions.6–8
LIMITATIONS OF THE STUDY
Because the assessments were taken on two dates only, the exact point in time from which only negligible changes of the movement patterns may be expected could not be determined. In addition, a minimal reduction of the interindividual variability after T2 cannot be completely excluded. For this reason, further studies are required to cover the whole adaptation phase with a higher number of assessments and shorter intervals between the assessments after fitting a new prosthesis.
A further limitation of this study is that findings and conclusions are, for the time being, applicable only to functionally similar knee joints and patients with higher functional levels. General conclusions for other prosthetic component groups or patients with lower functional levels will have to be investigated in separate studies.
The focus on the biomechanical parameters in the sagittal plane is based on the experiences reported in earlier studies.13–18,23 Compensatory motion patterns are to be expected particularly for ascending stairs in the coronal and transverse planes. This should be investigated in future studies by means of advanced marker models.
The results of the present study show that established unilateral transfemoral amputees may intuitively use the functions of a new prosthetic knee joint after only a few hours of adaptation if the motion pattern required for a given movement is similar to that of the previous prosthesis. This confirms that biomechanical tests to investigate the function of similar components are valuable even after such short period of acclimation.
For motion patterns that have yet to be learned with the new component, objective tests to detect important changes require a longer adaptation and learning phase. The results of the study suggest that a 3-month period may be appropriate.
The authors thank Kimberly Walsh, Anett Elsner, and David Reynolds for their valuable contributions to the preparation of this article.
1. Highsmith J, Kahle J, Bongiorni D, et al. Safety, energy efficiency and cost efficacy of the C-Leg for transfemoral amputees: a review of the literature. Prosthet Orthot Int
2010; 34: 362–377.
2. Johansson JL, Sherrill DM, Riley PO, et al. A clinical comparison of variable-damping and mechanically passive prosthetic knee devices. Am J Phys Med Rehabil
2005; 84: 563–575.
3. Kastner J, Nimmervoll R, Wagner P. Was kann das C-Leg?—Ganganalytischer Vergleich von C-Leg, 3R45 und 3R80. Med Orth Tech
1999; 119: 131–137.
4. Nimmervoll R, Kastner J, Kristen H. The C-Leg experience—a gait analysis
comparison with conventional prosthetic knee joints
2003; 54: 562–565.
5. Seymour R, Engbertson B, Kott K, et al. Comparison between C-Leg microprocessor-controlled prosthetic knee and non-microprocessor control prosthetic knees: a preliminary study of energy expenditure, obstacle course performance, and quality of life survey. Prosthet Orthot Int
2007; 31: 51–61.
6. Hafner B, Willingham L, Buell N, et al. Evaluation of function, performance and preference as transfemoral amputees transition from mechanical to microprocessor control of the prosthetic knee. Arch Phys Med Rehabil
2007; 88: 207–217.
7. Kahle J, Highsmith M, Hubard S. Comparison of non-microprocessor knee mechanism versus C-Leg on prosthesis evaluation questionnaire, stumbles, falls, walking tests, stair descent and knee preference. J Rehabil Res Dev
2008; 45: 1–14.
8. Kaufmann KR, Iverson B, Padgett D, et al. Do microprocessor-controlled knees work better? J Biomechanics
2006; 39: 70.
9. Klute G, Berge J, Orendurff M, et al. Prosthetic intervention on activity of lower-extremity amputees. Arch Phys Med Rehabil
2006; 87: 717–722.
10. Williams R, Turner A, Orendurff M, et al. Does having a computerized prosthetic knee influence cognitive performance during amputee walking? Arch Phys Med Rehabil
2006; 87: 989–994.
11. English R, Hubbard W, McElroy K. Establishment of consistent gait after fitting of new components. J Rehabil Res Dev
1995; 32 (1): 32–35.
12. Kaufmann KR, Levine JA, Brye RH, et al. Gait and balance of transfemoral amputees using passive mechanical and microprocessor controlled prosthetic knees. Gait Posture
2007; 26: 489–493.
13. Schmalz T, Blumentritt S, Jarasch R. Energy expenditure and biomechanical characteristics of lower limb amputee gait: The influence of prosthetic alignment and different prosthetic components. Gait Posture
2002; 16: 255–263.
14. Blumentritt S. Biomechanische Aspekte zur Indikation von Prothesenkniegelenken. Orthopädie-Technik
2004; 55: 508–521.
15. Bellmann M, Schmalz T, Ludwigs E, Blumentritt S. Immediate effects of a new microprocessor-controlled prosthetic knee joint: a comparative biomechanical evaluation. Arch Phys Med Rehabil
2012; 93 (3): 541–549.
16. Bellmann M, Schmalz T, Ludwigs E, Blumentritt S. Stair ascent with an innovative microprocessor-controlled exoprosthetic knee joint. Biomech Tech
2012; 57: 435–444.
17. Bellmann M, Schmalz T, Blumentritt S. Comparative biomechanical analysis of current microprocessor-controlled prosthetic knee joints
. Arch Phys Med Rehabil
2010; 91: 644–652.
18. Ludwigs E, Bellmann M, Schmalz T, Blumentritt S. Biomechanical differences between two exoprosthetic hip joint systems during level walking. Prosthet Orthot Int
2010; 34: 449–460.
19. Highsmith J, Kahle J, Shepard T, Kaufman K. The effects of the C-Leg knee prosthesis on sensory dependency and falls during sensory organization training. Technol Innov
2014; 15: 343–347.
20. Highsmith J, Kahle J, Miro M, et al. Percieved differences between the Genium and C-Leg microprocessor prosthetic knees in prosthetic-related function and quality of live. Technol Innov
2014; 15: 369–375.
21. Highsmith J, Kahle J, Lura J, et al. Stair ascent and ramp gait training with the Genium knee. Technol Innov
2014; 15: 349–358.
22. Nietert M. The Compromise Pivot Axis of the Knee Joint
. Aachen, Germany: Shaker; 2008.
23. Schmalz T, Blumentritt S, Altenburg B. Biomechanische Analyse des Schrägen—und Treppengehens mit aktuellen Kniepassteilen. Orthopädie-Technik
2006; 57: 682–693.
24. Woltring H. A Fortran package for generalized, cross-validatory spline smoothing and differentiation. Adv Eng Softw
1986; 8: 104–113.
25. Winter D. Biomechanics and Motor Control of Human Movement
. Waterloo, Canada: John Wiley & Sons; 2005.
26. Müller H, Blischke K. Motorisches Lernen. In: Schlicht W, Strauß B, eds. Grundlagen der Sportpsychologie
. Göttingen, Germany: Hogrefe; 2008.
27. Blischke K, Marschall F, Müller H, Daugs R. Augmented skill in motor skill acquisition. In: Van den Auweele Y, Bakker F, Biddle S, et al. eds. Psychology for Physical Educators
. Champaign, IL: Human Kinetics; 1999.