Osseointegrated Fixation: Solution for Transfemoral Amputation
Over the last 10 years, a few groups have developed innovative surgical methods of attachment of the prosthesis for transfemoral amputees that is based on direct skeletal anchorage. In this case, the socket is replaced by an osseointegrated fixation.1,2 One of the most advanced fixations includes an implant, an abutment, and a retaining bolt.3–5 The implant develops a firm biological bonding with the femur, named osseointegration.6–8 The abutment is connected to the implant, penetrating through the skin, to allow attachment of the external prosthesis.
Benefits of Osseointegrated Fixation
To date, this technique has been experienced by more than 100 transfemoral amputees worldwide mainly scattered in Scandinavia, the United Kingdom, Spain, and Australia.9 It has proved to be a successful alternative for amputees who experience complications in using conventional socket-type prostheses because of a short residual limb and soft tissue problems. This technique has contributed to a significant improvement in the quality of life of recipients.10–12 By definition, the biomechanical benefits of the fixation alone are limited, because it does not involve any articulated parts. Nevertheless, the fixation can improve sensory feedback, referred to as osseoperception,7 which might have indirect advantages for locomotion (e.g., foot placement and surface detection). The physical and prosthetic benefits are the most noticeable.11,13 The absence of a prosthetic socket can alleviate the skin problems and residual limb pain. It also enables greater hip range of motion and better sitting comfort compared with socket-type prostheses.14 The prosthetic leg can be attached to and detached from the fixation easily by simply turning a screw.
However, one of the burning issues for clinicians and funding bodies is to determine to what extent these prosthetic benefits are translated into an improvement in functional outcome. By definition, the term “functional outcome” corresponds to the capacity to undertake a wide range of tasks of daily living. Here, this term refers more precisely to the actual capacity of the amputee to use the prosthetic leg and to walk at self-selected pace.
So far, studies relying on questionnaires demonstrated that the fixation has increased the capacity of transfemoral amputees to improve their prosthetic activity and walking habits.11,13 Other studies described the loading of the fixation during standardized daily activities by using patterns, local extrema, and impulse.15,16 These indicators were essential for engineers designing the fixation, but they have limited clinical relevance in relation to functional outcome. One study looking at the load regime during real-world activities of daily living presented some functional outcome indicators, but it involved a single participant.17
Measurement of Functional Outcome
The functional outcome of transfemoral amputees can be assessed using a range of spatial and temporal gait characteristics. Some of the clinical indicators commonly acknowledged include the cadence, along with the duration of gait cycle, and support and swing phases.18–22
An overview of the resources and comprehensiveness of the output of equipments that are typically used to assess these four variables is shown in Figure 1.23–29 The most comprehensive and resource intensive assessment relies on a 3D motion analysis system synchronised with several force plates.30 Alternatively, studies demonstrated that other less resource intensive and equally accurate instruments can be used, including foot switch,31–35 pressure sensors,36 accelerometers and gyroscopes,32,37 instrumented mats, and walkways.30,31,38 Some instruments that can only be used in clinical settings measure a limited number of steps that are only partially representative of the true functional outcome. Other equipments that are portable enabled more realistic measurements.32,34,37 Previous studies focusing on the load applied on the fixation used a portable kinetic recording system based on a transducer and a wireless modem or a data logger.17,34,39 This system enabled the recording of an unlimited number of steps during various activities of daily living.15,16 As mentioned above, these studies focused mainly on presenting the load profile over time. Regrettably, the temporal variables of these data sets have yet to be presented.
Purpose and Objectives
The purpose of this study was to characterize the functional outcome of transfemoral amputees fitted with an osseointegrated fixation (TFA-OF) using key temporal gait variables. The objectives were as follows:
- To present the cadence, duration of gait cycle, and support and swing phases with an emphasis on the stride-to-stride and participant-to-participant variability, and
- To compare these temporal variables with normative data extracted from the literature focusing on transfemoral amputees fitted with a conventional socket (TFA-SO) and able-bodied participants.
The raw data used in the study have been published in Lee et al.,16 along with a detailed account of methodological aspects associated with the participants (e.g., profile of prosthesis), the apparatus, and the procedure. Consequently, only the most relevant information is presented here.
A total of three women and nine men with unilateral TFA-OF (47.50 ± 9.70 years, 1.78 ± 0.11 meter, and 84.27 ± 16.82 kg) participated in this study. Individual and group demographics are presented in Table 1 (Section A). Each participant was fully rehabilitated, fitted with the fixation for at least 1 year, able to walk 200 m independently, and weighed less than 110 kg to avoid overloading the transducer. All the participants were active and were classified as a K3 or K4 according to Functional Classification Levels.26 The research institution's human ethics committee approved this study. The participants provided informed written consent.
All participants walked with a prosthesis fitted with a transducer and their usual components,16 including:
- Hydraulic knee: polycadence responsive (i.e., 9 Total Knee®, Össur, Aliso Viejo, CA) and single axis cadence responsive [i.e., 1 Adaptive® (Blatchford & Sons Ltd, Basingstoke, UK), 1 C-Leg® (Otto Bock, Burlington, ON, Canada), 1 Mauch® Gaitmaster™ (Össur)];
- Foot: dynamic foot [i.e., 1 Mercury™ (Blatchford & Sons Ltd), 2 C-Walk® (Otto Bock), 1 Carbon Copy (Ohio Willow Wood, Mt. Sterling, OH), and 1 Flex-Foot® (Össur)], multiaxis foot [i.e., 2 Multiflex (Blatchford & Sons Ltd) and 3 TrueStep® (College Park Industries, Fraser, MI)], and single axis foot [i.e., 2 Total concept (Century XXII Innovations Inc., Jackson, MI)];
- Footwear (i.e., 5 sneakers, 4 sandals, and 3 leather shoes).
Also, the usual alignment was also replicated to insure ecological assessments.
The raw data were measured using a portable kinetic system with a sampling frequency of 200 Hz. It included a six-channel transducer (Model 45E15A; JR3, Inc., Woodland, CA) mounted between the knee and the fixation and a wireless transmitter (Ricochet Model 21062; Metricom, Inc., Los Gatos, CA).
Participants walked under supervision with the instrumented prosthesis for approximately 15 minutes to ensure confidence, safety, and comfort. Participants 1 and 2 performed six trials along a 20-m walkway at one site. The other participants performed approximately two trials along a 60-m walkway at another site. All participants walked at self-selected pace. Sufficient rest was given between trials to avoid fatigue.
The load data were processed by a customized Matlab program (Math works, Inc., Natick, MA) according to the following steps:
Step 1: Selection of Relevant Segment of Data
The first and the last strides recorded for each trial were discarded to ensure that the analysis only included data obtained when participants walked at a uniform pace.39
Step 2: Determination of Gait Events
The graph of the vertical force was used to manually detect the heel contact and toe-off points, with an accuracy of ±0.01 seconds as determined in previous study.40
Step 3: Calculation of Temporal Variables
The cadence expressed as the number of strides of the prosthetic leg per minute was calculated for each trial. The duration of a gait cycle was determined from heel contact to heel contact and expressed in seconds. The duration of support and swing phases were expressed in seconds and in percentage of the gait cycle.
Step 4: Characterization of Each Temporal Variable
The stride-to-stride analysis providing the variations for a given participant included the mean, standard deviation (SD), and coefficient of variation (COV), defined as the SD divided by the mean. The participant-to-participant analysis providing the variations between participants relied on the same descriptors and the median, minimum, and maximum values for the group.
Step 5: Comparative Analysis
The normative temporal variables were extracted from the literature focusing on TFA-SO41–49 and able-bodied participants.50–57 English publications up to 2008 were selected using mainstream search engines (e.g., PubMed, Medline, and Google Scholar) and combinations of keywords such as gait, walking, temporal variable, cadence, gait cycle, support phase, swing phase, transfemoral amputation, above-knee amputation, able-bodied, etc. The publications featuring literature review and/or meta-analyses were purposely excluded to avoid statistical compounding errors.18–22 The study presented by Frossard et al.17 was also excluded as it provided temporal variables for only one TFA-OF. One data set or more were extracted from each selected study to make sure that normative data matched as closely as possible the group of participants in terms of demographics (i.e., gender, age, height, and mass), procedure (i.e., self-selected walking speed), and fitting (i.e., hydraulic knee and foot). Each normative study reported one or more temporal variables.41–57 Some variables were recalculated based on raw data provided in the article (e.g., cadence in steps per minute/2 = cadence in strides per minute, gait cycle time = 120/cadence). The comparisons of the overall results for each group were based on the average. The significance of the differences for the study-to-study comparison was determined using a t-test with p < 0.0005 and p < 0.005 when the number of observations, the mean, and the SD were reported.
The total number of strides for each participant and for the group is presented in Section B of Table 1. It ranged from 32 to 63 strides, excluding participant 2 who performed only one trial. A total of 613 strides were assessed.
Characterization of Temporal Variables
The results of the stride-to-stride and participant-to-participant analyses are presented in Section C of Table 1. The individual COV of the cadence, duration of the gait cycle, and support and swing phases ranged from 0.007 to 0.042, 0.016 to 0.081, 0.026 to 0.089, and 0.028 to 0.112, respectively. The overall COV of the cadence, duration of the gait cycle, and support and swing phases were 0.076, 0.088, 0.089, and 0.125, respectively. The support and swing phases represented 57% ± 3% and 43% ± 3% of the gait cycle, respectively.
An overview of the temporal variables for the three groups is shown in Figure 2, plotting the cadence in relation to the duration of the gait cycle expressed in seconds, and duration of support and swing phases expressed in percentage of gait cycle. Tables 2 and 3 provide the raw data for demographics (Section A), number of samples (Section B), and normative temporal variables (Section C) for the studies focusing on TFA-SO and able bodied, respectively.
Comparison With Amputees Using Socket
The group of studies focusing on TFA-SO included nine data sets that corresponded to a total of 142 participants and 542 observations. Most of the participants were men.
This group of normative studies was 9% younger and 4% lighter. The study-to-study comparison was possible for only three data sets. One normative study was significantly younger. Two were significantly lighter. All the other comparisons, including the one related to the height, were not significant.
The cadence of this group was 2% slower. Only one of two possible comparisons was significantly quicker. The overall duration of the gait cycle and support and swing phases of this normative group were 3% and 6% shorter, and 1% longer, respectively. The duration of the gait cycle was significantly shorter for five studies and longer for one study of eight comparisons. The duration of the support phase was significantly shorter for four of five studies. The duration of the swing phase was significantly shorter for two studies and longer for two studies of seven comparisons.
Comparison With Able-Bodied Participants
The group of studies focusing on able bodied included 9 studies and 14 data sets, corresponding to a total of 258 participants and 1,603 observations. Only three data sets included female participants.
This group of normative studies was 16% younger and 7% lighter. The study-to-study comparison was possible for only two data sets of the same study. One data set involved significantly younger, smaller, and lighter women. The other one involved significantly younger men. Both groups had the same height.
The cadence of this group was 11% quicker. All eight possible study-to-study comparisons were significantly quicker. The overall duration of the gait cycle and support and swing phases of this normative group were 9%, 6%, and 13% shorter, respectively. The duration of the gait cycle was significantly shorter for seven studies and longer for one study of eight comparisons. The duration of the support phase was significantly shorter for four studies and longer for one study of five comparisons. The duration of the swing phase was significantly shorter for five studies of six comparisons.
Each participant presented a low stride-to-stride variability for the four temporal characteristics. This agrees with previous studies focusing on the magnitude and variability of the loading on fixation during walking and daily activities.15,16
Surprisingly, the participant-to-participant variability was lower than expected. For instance, previous study demonstrated that the COV of the peak forces applied on the anteroposterior, mediolateral, and long axes of the fixation during walking were, 0.523, 0.384, and 0.374, respectively.16 Temporal variables were more consistent than the loading characteristics may be because they are less sensitive to confounders such as the length of the residuum, prosthetic alignment, trunk position, etc.
The cadence, the duration of gait cycle, and support phases demonstrated that the functional outcome of TFA-OF was either comparable or better than TFA-SO in most studies. The outcome of the comparisons of the duration of the swing phase was more ambivalent. This was because the velocity of the leg during the swing depends less on functional outcome and more on the swing control of different types of knee friction systems (i.e., constant friction, variable friction, and hydraulic). Overall, the participants were less functional than able bodied. However, the most functional TFA-OF and the least functional able bodied were similar.
All combined, the clinical indicators observed in this study revealed that the fitting of an osseointegrated fixation has enabled this group of amputees to restore their locomotion with highly functional level. Further interpretations must be considered carefully giving the limitations of this study.
Group of Active Participants
The group of amputees represented approximately 15% of the existing population. The selection of the participants was as random as possible, giving the pool and the location worldwide. Nonetheless, the design of the study itself was slightly biased toward recruiting enable and active participants, similar to any other studies based on several trials of walking. Consequently, finding good functional outcomes was to be expected. Thus, the tangible contribution of this study was to determine to which extent functional outcome was satisfactory.
Comparison With Normative Studies
In principle, the meta-analysis of normative data extracted from the literature enabled the use of large data sets that have already been validated. However, no studies reported complete demographic and temporal variable data sets. Only 30% of the studies in each group presented three complete temporal variable data sets. Furthermore, a true comparison might be compromised because of the possible multiple use of the same population across several studies from the same authors. This created a potential redundancy in the statistical analysis as the same individual might have been considered several times. Finally, comparisons might be interpreted with care because confounders matched only partially. The height, which is one of the critical confounders, was the same for the three groups. The age and the mass presented small but significant differences. This study included a large female population. More importantly, there were variations in construction of prostheses in terms of components, particularly the knees, and alignments. Other confounders associated with instruments (e.g., foot switch, force plate, and motion analysis), procedure (e.g., accommodation time with experimental leg), and inclusion/exclusion criteria of population (e.g., cause of amputation, level of activity) were not considered.18
These limitations do not impinge on three main outcomes of this study. The results presented here can be used to benchmark other cohorts against the most active TFA-OF. The low variability means that temporal variables can be used as default input in generic algorithms using templates of patterns based on timing and magnitude of loading to recognize activities of daily living and to detect falls. Finally, this study confirmed that the portable kinetic system used is a suitable instrument to provide not only engineering (i.e., patterns, local extrema, and impulse) but also clinical (i.e., temporal variables) insights into the fitting and usage of the prosthetic leg. The seamlessness of the system enabled the recording of a high number of strides. For instance, this study alone collected approximately 12% more strides than all nine studies in the groups of TFA-SO combined.
The instrument presented here will facilitate longitudinal studies of temporal characteristics of standardized and real-world activities of daily living for a larger cohort of TFA-OF. This will provide a better understanding of the participant-to-participant and activity-to-activity variability. Comparisons of the results from current instruments and the portable kinetic system were outside the scope of this study. However, the possibilities for cross-sectional studies are endless, particularly for the ones allowing reciprocal validation of these instruments (e.g., accuracy), recording complementary clinical indicators such as the temporal variables for the sound leg (e.g., duration of single and double support) and the spatial variables (e.g., step and stride length, walking base), as well as comparisons between prostheses constructions (e.g., hydraulic knee vs. constant friction).
Both longitudinal and cross-sectional studies will be essential to further establish the functional outcome of TFA-OF in terms of usage of the prosthesis and level of activity. Furthermore, such studies will improve basic knowledge in the areas of rehabilitation, design of components, and fitting of prostheses.
This study provided the temporal gait characteristics for a group of 12 TFA-OF. This was the first attempt to establish to what extent the benefits of this innovative method of attachment of the prosthesis are translated into functional outcome and, more particularly, walking ability. The results indicated that the fixation enables this group to walk as well or better than other amputees fitted with a socket, although this statement must be understood within the intrinsic limitations of temporal variables and comparisons with data from the literature. Consequently, further longitudinal and cross-sectional studies would be required to confirm these results.
In conclusion, the results presented here are a stepping stone in assessment of true functional outcome of transfemoral amputees fitted with a fixation. However, this study provided key information to clinicians facing the challenge to restore the locomotion of lower limb amputees in the framework of an evidence-based practice.
1.Aschoff H, Grundei H, eds. The Endo-Exo-Femurprosthesis: A New Concept of Prosthetic Rehabilitation Engineering Following Thigh-Amputation—Some Cases and Early Results
. Hong Kong: International Society for Prosthetics and Orthotics 11th World Congress; 2004.
2.Staubach K, Grundei H. The first osseointegration
percutaneous anchor for an exoprosthesis, for routine use in above-knee amputees [in German]. Biomed Tech (Berl)
3.Brånemark R, Brånemark P-I, Rydevik B, Myers RR. Osseointegration
in skeletal reconstruction and rehabilitation: a review. J Rehabil Res Dev
4.Robinson KP, Brånemark R, Ward D. Future developments: osseointegration
in transfemoral amputees. In: Smith DG, Michael JW, Bowker JH, eds. Atlas of Amputations and Limb Deficiencies: Surgical, Prosthetic and Rehabilitation Principles.
3rd ed. Chapter 53. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2005:673–681.
5.Ward D, Robinson KP. Osseointegration
for the skeletal fixation of limb prostheses in amputation at the trans-femoral level. In: Brånemark P-I, ed. The Osseointegration Book–From Calvarium to Calcaneus
. Chapter 24. Quintessenz Verlag: GmbH; 2005:463–475.
6.Pitkin M, Raykhtsaum G, Galibin OV, Protasov M. Skin and bone integrated prosthetic pylon: a pilot animal study. J Rehabil Res Dev
7.Brånemark R, Ohrnel LO, Nilsson P, Thomsen P. Biomechanical characterization of osseointegration
during healing: an experimental in vivo study in the rat. Biomaterials
8.Brånemark R, Skalak R. An in-vivo method for biomechanical characterization of bone-anchored implant. Med Eng Phys
9.Hagberg K, Brånemark R. One hundred patients treated with osseointegrated transfemoral amputation
prostheses—the rehabilitation perspective. J Rehabil Res Dev
10.Ax E, ed. The Personal Dimension of Innovation: Making a Personal Choice to Improve Quality of Life With Innovation
. Hong Kong: International Society for Prosthetics and Orthotics 11th World Congress; 2004.
11.Hagberg K, Branemark R, Gunterberg B, Rydevik B. Osseointegrated transfemoral amputation
prostheses: prospective results of general and condition-specific quality of life in 18 patients at 2-year follow-up. Prosthet Orthot Int
12.Webster JB, Chou TP, Kenly M, et al. Perceptions and acceptance of osseointegration
among individuals with lower limb amputations: a prospective survey study. J Prosthet Orthot
13.Sullivan J, Uden M, Robinson K, Sooriakumaran S. Rehabilitation of the trans-femoral amputee with an osseointegrated prosthesis: the United Kingdom experience. Prosthet Orthot Int
14.Hagberg K, Haggstrom E, Uden M, Brånemark R. Socket versus bone-anchored trans-femoral prostheses: hip range of motion and sitting comfort. Prosthet Orthot Int
15.Lee W, Frossard L, Hagberg K, et al. Kinetics analysis of transfemoral amputees fitted with osseointegrated fixation performing common activities of daily living. Clin Biomech (Bristol, Avon)
16.Lee WC, Frossard LA, Hagberg K, et al. Magnitude and variability of loading on the osseointegrated implant of transfemoral amputees during walking. Med Eng Phys
17.Frossard L, Stevenson N, Smeathers J, et al. Monitoring of the load regime applied on the osseointegrated fixation of a transfemoral amputee: a tool for evidence-based practice. Prosthet Orthot Int
18.Gard S. Use of quantitative gait
analysis for the evaluation of prosthetic walking performance. J Prosthet Orthot
19.Skinner HB, Effeney DJ. Gait
analysis in amputees. Am J Phys Med
20.Whittle M. Gait Analysis
. 2nd ed. Oxford, UK: Butterworth-Heinemann; 1996.
21.Perry J. Stride analysis. Gait Analysis: Normal and Pathological Function.
Chapter 20. Thorofare, NJ: SLACK, Inc; 1992:431–441.
22.Whittle MW. Clinical gait
analysis: a review. Hum Mov Sci
23.Gailey RS, Roach KE, Applegate EB, et al. The amputee mobility predictor: an instrument to assess determinants of the lower-limb amputee's ability to ambulate. Arch Phys Med Rehabil
24.Miller L, McCay JA. Summary and conclusions from the academy's sixth state-of-the-science conference on lower limb prosthetic outcome measures. J Prosthet Orthot
25.Condie E, Scott H, Treweek S. Lower limb prosthetic outcome measures: a review of the literature 1995 to 2005. J Prosthet Orthot
26.Gailey RS. Predictive outcome measures versus functional outcome
measures in the lower limb amputee. J Prosthet Orthot
27.Gauthier-Gagnon C, Grisé M-C. Tools to measure outcome of people with a lower limb amputation: update on the PPA and LCI. J Prosthet Orthot
28.Boone DA, Coleman KL. Use of the prosthesis evaluation questionnaire (PEQ). J Prosthet Orthot
29.Coleman KL, Smith DG, Boone DA, et al. Step activity monitor: long-term continuous recording of ambulatory function. J Rehabil Res Dev
30.Ghoussayni S, Stevens C, Durham S, Ewins D. Assessment and validation of a simple automated method for the detection of gait
events and intervals. Gait Posture
31.Barker S, Craik R, Freedman W, et al. Accuracy, reliability, and validity of a spatiotemporal gait
analysis system. Med Eng Phys
32.Aminian K, Najafi B, Büla C, et al. Spatio-temporal parameters of gait
measured by an ambulatory system using miniature gyroscopes. J Biomech
33.Kyriazis V, Rigas C, Xenakis T. A portable system for the measurement of the temporal parameters of gait
. Prosthet Orthot Int
34.Holden J, Fernie GR, Soto M. An assessment of a system to monitor the activity of patients in a rehabilitation programme. Prosthet Orthot Int
35.Hausdorff JM, Ladin Z, Wei JY. Footswitch system for measurement of the temporal parameters of gait
. J Biomech
36.Barnett S, Cunningham JL, West S. A comparison of vertical force and temporal parameters produced by an in-shoe pressure measuring system and a force platform. Clin Biomech
37.Zijlstra W. Assessment of spatio-temporal parameters during unconstrained walking. Eur J Appl Physiol
38.Lloyd DG, Svensson NL. A technique for the measurement of cadence using walkway vibrations. J Biomech
39.Frossard L, Beck J, Dillon M, et al. Development and preliminary testing of a device for the direct measurement of forces and moments in the prosthetic limb of transfemoral amputees during activities of daily living. J Prosthet Orthot
40.Dumas R, Cheze L, Frossard L. Loading applied on prosthetic knee of transfemoral amputee: comparison of inverse dynamics and direct measurements. Gait Posture
41.Jaegers SM, Arendzen JH, de Jongh HJ. Prosthetic gait
of unilateral transfemoral amputees: a kinematic study. Arch Phys Med Rehabil
42.Macfarlane PA, Nielsen DH, Shurr DG. Mechanical gait
analysis of transfemoral amputees: SACH foot versus the flex-foot. JPO J Prosthet Orthot
43.Hale SA. Analysis of the swing phase dynamics and muscular effort of the above-knee amputee for varying prosthetic shank loads. Prosthet Orthot Int
44.Boonstra AM, Schrama W, Eisma WH, et al. Gait
analysis of transfemoral amputee patients using prostheses with two different knee joints. Arch Phys Med Rehabil
45.Murray MP, Sepic SB, Gardner GM, Mollinger LA. Gait
pattern of above-knee amputees using constant friction knee components. Bull Prosthet Res
46.Zuniga E, Leavitt L, Calvert J, et al. Gait
patterns in above-knee amputees. Arch Phys Med Rehabil
47.James U, Oberg K. Prosthetic gait
pattern in unilateral above-knee amputees. Scand J Rehabil Med
48.Godfrey C, Jousse A, Brett R, Butler J. A comparison of some gait
characteristics with six knee joints. Orthot Prosthet
49.Murray M, Mollinger L, Sepic S, Gardner G. Gait
patterns in above-knee amputee patients: hydraulic swing control vs constant-friction knee components. Arch Phys Med Rehabil
50.Murray M. Gait
as total pattern of movement. Am J Phys Med
51.Andriacchi TP, Ogle JA, Galante JO. Walking speed as a basis for normal and abnormal gait
measurements. J Biomech
52.Murray M, Drought A, Kory R. Walking patterns of normal men. J Bone Joint Surg Am.
53.Murray M, Kory R, Clarkson B, Sepic S. Comparison of free and fast speed walking patterns of normal men. Am J Phys Med
54.Kadaba M, Ramakrishnan H, Wootten M, et al. Repeatability of kinematic, kinetic, and electromyographic data in normal adult gait
. J Orthop Res
55.Oberg T, Karsznia A. Basic gait
parameters: reference data for normal subjects, 10–79 years of age. J Rehabil Res Dev
56.Allard P, Lachance R, Aissaoui RS, et al. Able-bodied gait
in men and women. In:Allard AC, Lundberg A, Vaughan C, eds. Three-Dimensional Analysis of Human Locomotion
. Chapter 15. New York, NY: Jonh Wiley & Sons; 1997:307–334.
57.Kadaba M, Ramakrishnan H, Wootten M. Measurement of lower extremity kinematics during level walking. J Orthop Res