The majority of combat-related injuries sustained during Operations Iraqi Freedom and Enduring Freedom involve the extremities . These extremity injuries are often the result of explosive mechanisms (52%), gunshots (16%), or mortar attacks (9%) . Advances in surgical procedures  and rehabilitation [7, 31, 32] have improved the ability to salvage the limb and restore function after severe injury. Limb salvage, or reconstruction, is a viable treatment option for many patients with severe limb trauma who would otherwise undergo amputation . However, many patients still are unable to return to full functional capacity as a result of muscle loss , instability, stiffness, chronic pain, and peripheral nervous system injury.
Ankle-foot orthoses (AFOs) commonly are prescribed to provide mechanical support to the salvaged limb during walking and other functional tasks. The external support provided can enhance performance outcomes [4, 14-16, 33, 40] and stability  by counteracting joint torque  and improving proprioception  to reduce injury . Although the majority of AFO research is conducted using populations with myelomeningocele, spastic diplegia, hemiparesis, and multiple sclerosis, a common feature of AFO use in these populations and those with limb salvage is plantarflexor weakness. Gait is hindered by limited plantarflexor power [26, 27] for which the hip generally compensates [24, 26]. The resulting gait is mechanically inefficient [8, 22] and leads to elevated energy cost [29, 39]. Most passive-dynamic AFOs help compensate by functioning as a spring that stores energy when initially deformed in midstance and returns energy at the end of stance [2, 12, 42]. The stiffness of a dynamic AFO can be optimized to alleviate gait-related problems [4, 18, 36] because it determines the extent to which the AFO maintains the ankle in a neutral position, provides mediolateral stability, and aids propulsion through energy storage and return mechanisms [25, 36].
The biomechanical effects of dynamic AFO use after lower extremity trauma have not been widely reported. Patzkowski et al.  was the first to compare three different dynamic AFOs during performance tasks in military patients with limb salvage. These AFOs were the posterior leaf spring, Blue Rocker (Allard USA Inc, Rockaway, NJ, USA), and a new custom dynamic AFO available to wounded warriors called the Intrepid Dynamic Exoskeletal Orthosis (IDEO, patent pending #20120271214). The IDEO offered functional and performance improvements over the other AFOs and several patients decided against limb amputation after rehabilitation with the IDEO. The IDEO mechanically compensates for insufficient ankle function (for example, the semirigid nature of the IDEO can compensate for a completely flaccid ankle) and has enabled wounded service members to return to high levels of physical activity . The stiffness of the IDEO, which affects its energy storage and return capabilities, is expected to play a large role in device functionality, but this has not been studied. In addition, despite functional and performance improvements in this patient population using the IDEO compared with other devices, it remains unknown if this dynamic AFO can restore normal gait mechanics.
Therefore, the overall purpose of this study was to determine the effect of dynamic AFO stiffness on lower extremity kinematics and kinetics in injured service members who had undergone lower limb salvage. Specifically, the first aim was to determine if AFO stiffness affected biomechanical parameters of walking such as joint angles, moments, and powers. The second aim was to determine if a given stiffness could normalize these gait parameters to control subjects.
Patients and Methods
This study used a repeated-measures, controlled study design to address the two aims. Thirteen male patients with traumatic, unilateral lower limb reconstruction gave written informed consent to participate in the study. The number of subjects was determined from preliminary data using a power of 0.80, an alpha of 0.05, and the ability to detect a difference of 0.56 W/kg in ankle power generation at pushoff . Mechanisms of injuries included motor vehicle accidents, gunshot wounds, and blasts. As a result of these injuries, patients were frequent or constant users of a custom IDEO ankle foot orthosis (Fig. 1). All patients were under the care of the same certified orthotist, had sustained functional limitations associated with traumatic lower limb injury, were ambulatory without assistive devices other than an AFO, and were capable of completing the study protocol. Inclusion criteria involved the ability to walk, run, perform agility-based movements during running, and climb stairs in the IDEO. Thirteen men with no history of lower extremity trauma served as a control group to provide normative walking data and walked without external or in-shoe orthoses. Subject groups were not different in height (p = 0.682) or body mass (p = 0.787) but the control group was, on average, 6 years younger (p = 0.005) (Table 1).
The experimental setup consisted of a 26-camera motion capture system (120 Hz; Motion Analysis Corp, Santa Rosa, CA, USA) with five centrally located force platforms in tandem along a walkway (1200 Hz; AMTI, Watertown, MA, USA). Fifty-seven retroreflective markers were secured to anatomical landmarks and segments of the upper and lower extremities, head, trunk, and pelvis. Rigid plates of markers were secured to the thighs and lower legs for tracking purposes  (Fig. 2). A digitization pointer consisting of four markers was used to identify 20 anatomical landmarks in relation to marker clusters (C-Motion, Inc, Germantown, MD, USA).
Subjects participated in three overground walking sessions on separate days. Stiffness of the IDEO was modified by altering the stiffness of the posterior strut component. Custom struts were manufactured specifically for study purposes to allow the assessment of three conditions: (1) nominal (clinically prescribed stiffness as fit by the prosthetist/orthotist); (2) compliant (20% more compliant than the nominal strut); and (3) stiff (20% stiffer than the nominal strut). Clinical prescription of strut stiffness was based on the patient’s available range of motion (ROM), activity level, types of activities performed, body mass, and active duty status (ie, indicating that they would carry heavy loads on a regular basis). All struts were initially designed in SolidWorks (Waltham, MA, USA) and constructed from Unfilled Nylon 11 powder (PA D80-ST; Advanced Laser Materials, Temple, TX, USA) using a selective laser sintering technique previously described [13, 35]. Mechanical testing performed before biomechanical testing ensured the struts were within 5% of their intended stiffness. Mechanical testing indicated that the nominal struts ranged in stiffness from 50 kgf/mm to 105 kgf/mm with mean (SD) values of 77 (19) kgf/mm. The constructed struts were affixed to the posterior aspect of the IDEO’s footplate and connected the rigid carbon fiber footplate to the upper tibial cuff. Strips of lead tape (ClubmakerTM; Golfsmith, Austin, TX, USA) were added along the entire lengths of the nominal and compliant struts to match the mass of the stiff strut. Then, patient participants wore their IDEO for at least 30 minutes before data collection. The footwear worn by patients was standardized among the three sessions but shoe makes and models were not standardized between subjects or between groups. Patients were not informed which strut they were wearing, and the order of the conditions was randomized. Patient preference for strut stiffness was recorded after the final testing session at which time subjects indicated which strut they preferred for daily use.
Three-dimensional marker and analog data were recorded as subjects walked across the force platforms at a standardized velocity. This velocity (± 5%) was calculated from the forward progression of a marker on the seventh cervical vertebrae and corresponded to a dimensionless Froude number of 0.16 . Self-selected walking velocity was also recorded as a descriptive characteristic. Five strides from the IDEO side were analyzed (Visual3D™, Version 4.96; C-Motion, Inc, Germantown, MD, USA). Marker and analog data were interpolated using a cubic spline and filtered using a fourth-order Butterworth low-pass filter with cutoff frequencies of 6 and 50 Hz, respectively.
A 15-segment, full body model was created in Visual3D and used in subsequent analyses . Sagittal plane ankle, knee, and hip angles were calculated. To control for any subtle alignment changes when struts changed, ankle angles were scaled to the value at 75% of swing, when the ankle maintained a relatively fixed position within the IDEO. Internal joint moments were calculated from a standard inverse dynamics approach and then resolved into their sagittal plane component and scaled to body mass. Lastly, joint powers were scaled to body mass and ground reaction forces were scaled to body weight. All dependent measures were normalized to 101 data points and represented as percent of stride.
Peak ankle, knee, and hip angles, internal moments and powers, and ground reaction forces were calculated (Matlab Version 7.14; The Mathworks, Inc, Natick, MA, USA) and included in the statistical analysis (Version 19; SPSS Inc, Chicago, IL, USA). To address the first aim, comparisons were made among the three stiffnesses of the orthoses tested using a one-way analysis of variance (ANOVA) and Huynh-Feldt corrections. Post hoc paired t-tests with Bonferroni-Holm correction factors were used to identify differences between stiffnesses. To address the second aim, comparisons between AFO users who had undergone limb salvage surgery after major lower extremity trauma and control subjects were tested using a one-way ANOVA with Dunnett’s post hoc tests. The unadjusted criterion for statistical significance was set at p < 0.05.
Aim 1: Comparisons Among Stiffnesses
Subject preference for the three strut stiffnesses varied. Three subjects preferred the stiff strut, five preferred the nominal, three preferred the compliant, one preferred nominal and stiff equally, and one could not discern any difference. Temporal spatial parameters of gait, including self-selected walking velocity, were not different among strut conditions (Table 2). Dynamic AFO stiffness influenced walking mechanics not at the ankle, but at the knee. Use of a 20% more compliant strut than prescribed resulted in a 20% decrease in stance phase knee flexion relative to the nominal strut (p = 0.003) and a 26% decrease relative to the stiff strut (p = 0.001) (Fig. 3; Table 2). Moreover, the peak internal knee extensor moment during stance was greater in the stiff strut compared with the nominal (p = 0.004) (Fig. 4). Stiffness did not affect peak joint powers (Fig. 5) or peak ground reaction forces (Fig. 6).
Aim 2: Comparisons Between Groups
Patients who used semirigid AFOs after limb reconstruction surgery exhibited gait deviations relative to able-bodied control subjects. None of the stiffnesses entirely normalized ankle, knee, or hip kinematics to controls (Fig. 3). Ankle and knee motion were more limited in the patient group and patients wearing the semirigid IDEO had, on average, 16 fewer degrees of available ankle ROM (p < 0.001) and 10 fewer degrees of knee ROM across the gait cycle than controls (p < 0.001) (Table 2). The hip compensated for some of the limitations in the other joints by increasing peak flexion during stance an average of 26% relative to controls (p < 0.013). Ankle power absorption in midstance and generation in late stance were attenuated in the patient group (Fig. 5). On average, peak ankle power absorption was 28% less than controls (p < 0.037) and the stiff strut reduced knee power generation at initial loading (p = 0.017). An average 64% reduction in ankle power generation relative to controls (p < 0.001) contributed to a 22% average decrease in the peak propulsive force relative to controls (p < 0.001) (Fig. 6).
The stiffness of a dynamic AFO plays a role in how it stores and returns energy during the gait cycle. However, it was unknown if certain stiffness parameters for AFOs worn by individuals with traumatic lower limb salvage could begin to normalize ankle, knee, and hip mechanics to able-bodied individuals. The primary purpose of this study was to determine the effect of AFO stiffness on biomechanical parameters of walking in patients with lower limb reconstruction. The secondary purpose was to compare gait parameters with control subjects.
This study had several limitations. Some of the gait limitations found in the IDEO condition relative to controls may have been influenced by the use of a very supportive AFO or may have resulted from the injury and/or surgical procedures themselves. Because it was not possible for many subjects to complete a gait analysis without their IDEO, no conclusions can be made regarding the specific effect of IDEO use compared with an unbraced condition. These comparisons are potentially interesting but may not have real-world relevance because patients depended on its external support for mobility. In addition, although the injuries sustained by the subjects were heterogeneous, between-subject variability in the patient group was not beyond that of noninjured individuals. Thus, although injuries, and possibly also ankle strength and ROM, differed within the patient group, these factors did not likely affect gait mechanics or the overall interpretation of the results of the study. Also, self-selected walking velocity is an indicator of recovery during rehabilitation after lower limb trauma  and velocities were not significantly different between the patient and control groups. Therefore, the patients may have attained a level of functional recovery where relevant comparisons could be made to control subjects during walking. The final limitation is that high-impact activities, where stiffer AFOs may offer more appropriate energy storage and return and reduce the risk of mechanical failure, were not tested.
Strut stiffness had some effects on gait. An increase in AFO compliance increased the stiffness at the knee because the knee underwent less flexion when the compliant strut was worn relative to the nominal and stiff. In agreement with Kobayashi et al. , decreased resistance about the ankle resulted in less stance-phase knee flexion. These results are also in agreement with footwear literature that reports less knee flexion in soft midsole (ie, more compliant) running shoes . In theory, an AFO with spring-like material properties that stores energy during midstance and returns the maximum amount of energy during late stance/preswing should be most beneficial to the user [3, 12, 42]. However, Bregman et al.  undertook a simulation approach to investigate how different stiffnesses affected walking biomechanics and found that the optimal (least metabolic cost) stiffness occurred when hip compensations were minimized. In the present study, no differences were found in hip kinematics or kinetics across strut stiffnesses potentially indicating that use of any of the three stiffnesses would result in similar metabolic costs. Overall, the biomechanical differences among struts were relatively small and a 40% range in stiffness did not drastically affect gait mechanics. In clinical practice, it is not possible to construct many AFOs to test optimal stiffness and clinicians and orthotists must make educated decisions based on available literature, prior experience, and patient needs. Although researchers are attempting to optimize dynamic AFO mechanical properties using modeling and simulation [6, 9], the results of this experimental study show that a range of stiffnesses may be equally beneficial to walking biomechanics.
Wearing a custom dynamic AFO after surgical treatment of limb trauma did not normalize gait to that of noninjured individuals and gait deficiencies remained across the lower extremity. Although knee ROM was lower in the patient group, the only differences in knee angle from control subjects occurred in swing and limb salvage patients did not adopt the stiff-kneed gait shown in previous reports . The patients also walked with greater hip flexion across the gait cycle. The stiff design of the IDEO along with impaired strength in the patients resulted in limited ankle plantarflexion. Many of the gait deviations such as decreased ankle motion and power were expected and are in agreement with previous literature on AFO use [5, 21]. To provide the support and stability necessary to achieve walking gait, some sacrifices in normative biomechanics must be made. Although no comparisons can be made to gait without the use of the IDEO, it is reasonable to expect that its use assisted and improved overall gait given the inability of some patients to walk without the device and the clinician and patient preference to use the device.
A more compliant AFO increases the stiffness at the knee through less knee flexion. However, AFO stiffness had few other effects on gait mechanics or on subjective preferences in patients with lower limb reconstructions. Patients may have readily adapted to the 40% stiffness range because walking did not stress the capabilities of the IDEO to the extent other dynamic activities such as running or jumping may have. Use of a semi-rigid, dynamic AFO inherently reduced the ROM and power capabilities at the ankle relative to controls and compensations at more proximal joints such as the knee resulted. Although none of the stiffness conditions restored all biomechanical gait parameters to those of control subjects, self-selected walking velocity was restored and previous reports have shown greater performance benefits with the IDEO compared with commercially available designs . For walking, if a range of dynamic AFO stiffnesses may be appropriately prescribed, this may reduce the burden on the orthotist to experimentally test numerous designs to find the best stiffness characteristics for individual patients. Selecting a stiffer dynamic AFO may be preferred for individuals who engage in high-impact activities (eg, running, jumping) to offer appropriate energy return with a lower risk of mechanical failure.
We thank Harmony Choi, Derek Haight, Jennifer Aldridge Whitehead, Kelly Rodriguez, Dr Deanna Gates, and Dr Richard Neptune for their contributions to this project.
1. Baker, PA. and Hewison, SR. Gait recovery pattern of unilateral lower limb amputees during rehabilitation. Prosthet Orthot Int.
1990; 14: 80-84.
2. Bartonek, A., Eriksson, M. and Gutierrez-Farewik, EM. Effects of carbon fibre spring orthoses on gait in ambulatory children with motor disorders and plantarflexor weakness. Dev Med Child Neurol.
2007; 49: 615-620. 10.1111/j.1469-8749.2007.00615.x
3. Bartonek, A., Eriksson, M. and Gutierrez-Farewik, EM. A new carbon fibre spring orthosis for children with plantarflexor weakness. Gait Posture.
2007; 25: 652-656. 10.1016/j.gaitpost.2006.07.013
4. Bregman, DJ., Groot, V., Diggele, P., Meulman, H., Houdijk, H. and Harlaar, J. Polypropylene ankle foot orthoses to overcome drop-foot gait in central neurological patients: a mechanical and functional evaluation. Prosthet Orthot Int.
2010; 34: 293-304. 10.3109/03093646.2010.495969
5. Bregman, DJ., Harlaar, J., Meskers, CG. and Groot, V. Spring-like Ankle Foot Orthoses reduce the energy cost of walking by taking over ankle work. Gait Posture.
2012; 35: 148-153. 10.1016/j.gaitpost.2011.08.026
6. Bregman, DJJ., Krogt, MM., Groot, V., Harlaar, J., Wisse, M. and Collins, SH. The effect of ankle foot orthosis stiffness on the energy cost of walking: a simulation study. Clin Biomech.
2011; 26: 955-961. 10.1016/j.clinbiomech.2011.05.007
7. Castillo, RC., MacKenzie, EJ., Archer, KR., Bosse, MJ. and Webb, LX. Evidence of beneficial effect of physical therapy after lower-extremity trauma. Arch Phys Med Rehabil.
2008; 89: 1873-1879. 10.1016/j.apmr.2008.01.032
8. Collins SH, Kuo AD. Recycling energy to restore impaired ankle function during human walking. Plos One.
9. Crabtree, CA. and Higginson, JS. Modeling neuromuscular effects of ankle foot orthoses (AFOs) in computer simulations of gait. Gait Posture.
2009; 29: 65-70. 10.1016/j.gaitpost.2008.06.004
10. Visser, E., Mulder, T., Schreuder, HWB., Veth, RPH. and Duysens, J. Gait and electromyographic analysis of patients recovering after limb-saving surgery. Clin Biomech.
2000; 15: 592-599. 10.1016/S0268-0033(00)00021-8
11. Dempster W. Space Requirements of the Seated Operator. WADC Technical Report (TR-55-159)
. Dayton, OH, USA: Wright Patterson Air Force Base; 1955.
12. Desloovere, K., Molenaers, G., Gestel, L., Huenaerts, C., Campenhout, A., Callewaert, B., Walle, P. and Seyler, J. How can push-off be preserved during use of an ankle foot orthosis in children with hemiplegia? A prospective controlled study. Gait Posture.
2006; 24: 142-151. 10.1016/j.gaitpost.2006.08.003
13. Faustini, MC., Neptune, RR., Crawford, RH. and Stanhope, SJ. Manufacture of passive dynamic ankle-foot orthoses using selective laser sintering. IEEE Trans Biomed Eng.
2008; 55: 784-790. 10.1109/TBME.2007.912638
14. Feuerbach, JW. and Grabiner, MD. Effect of the aircast on unilateral postural control—amplitude and frequency variables. J Orthop Sport Phys.
1993; 17: 149-154. 10.2519/jospt.19126.96.36.199
15. Greene, TA. and Hillman, SK. Comparison of support provided by a semirigid orthosis and adhesive ankle taping before, during, and after exercise. Am J Sport Med.
1990; 18: 498-506. 10.1177/036354659001800509
16. Greene, TA. and Wight, CR. A comparative support evaluation of three ankle orthoses before, during, and after exercise. J Orthop Sports Phys Ther.
1990; 11: 453-466. 10.2519/jospt.19188.8.131.523
17. Grogan, BF. and Hsu, JR. Skeletal Trauma Research Consortium. Volumetric muscle loss. J Am Acad Orthop Surg.
2011; 19: S35-S37.
18. Harlaar, J., Brehm, M., Becher, JG., Bregman, DJJ., Buurke, J., Holtkamp, F., Groot, V. and Nollet, F. Studies examining the efficacy of Ankle Foot Orthoses should report activity level and mechanical evidence. Prosthet Orthot Int.
2010; 34: 327-335. 10.3109/03093646.2010.504977
19. Hartsell, HD. The effects of external bracing on joint position sense awareness for the chronically unstable ankle. J Sport Rehabil.
2000; 9: 279-289.
20. Hartsell, HD. and Spaulding, SJ. Effectiveness of external orthotic support on passive soft tissue resistance of the chronically unstable ankle. Foot Ankle Int.
1997; 18: 144-150. 10.1177/107110079701800306
21. Kobayashi, T., Leung, AK., Akazawa, Y. and Hutchins, SW. The effect of varying the plantarflexion resistance of an ankle-foot orthosis on knee joint kinematics in patients with stroke. Gait Posture.
2013; 37: 457-459. 10.1016/j.gaitpost.2012.07.028
22. Kuo, AD. Energetics of actively powered locomotion using the simplest walking model. J Biomech Eng.
2002; 124: 113-120. 10.1115/1.1427703
23. Lehmann, JF., Condon, SM., Delateur, BJ. and Price, R. Gait abnormalities in peroneal nerve paralysis and their corrections by orthoses—a biomechanical study. Arch Phys Med Rehabil.
1986; 67: 380-386.
24. Lewis, CL. and Ferris, DP. Walking with increased ankle pushoff decreases hip muscle moments. J Biomech.
2008;41:2082-2089. 10.1016/j.jbiomech.2008.05.013. 2562040
25. Miyazaki, S., Yamamoto, S. and Kubota, T. Effect of ankle-foot orthosis on active ankle moment in patients with hemiparesis. Med Biol Eng Comput.
1997; 35: 381-385. 10.1007/BF02534094
26. Nadeau, S., Gravel, D., Arsenault, AB. and Bourbonnais, D. Plantarflexor weakness as a limiting factor of gait speed in stroke subjects and the compensating role of hip flexors. Clin Biomech.
1999; 14: 125-135. 10.1016/S0268-0033(98)00062-X
27. Neptune, RR., Kautz, SA. and Zajac, FE. Contributions of the individual ankle plantar flexors to support, forward progression and swing initiation during walking. J Biomech.
2001; 34: 1387-1398. 10.1016/S0021-9290(01)00105-1
28. Nigg, BM., Baltich, J., Maurer, C. and Federolf, P. Shoe midsole hardness, sex and age effects on lower extremity kinematics during running. J Biomech.
2012; 45: 1692-1697. 10.1016/j.jbiomech.2012.03.027
29. Olgiati, R., Burgunder, JM. and Mumenthaler, M. Increased energy-cost of walking in multiple-sclerosis—effect of spasticity, ataxia, and weakness. Arch Phys Med Rehabil.
1988; 69: 846-849.
30. Owens, BD., Kragh, JF., Macaitis, J., Svoboda, SJ. and Wenke, JC. Characterization of extremity wounds in operation Iraqi freedom and operation enduring freedom. J Orthop Trauma.
2007; 21: 254-257. 10.1097/BOT.0b013e31802f78fb
31. Owens, JG. Physical therapy of the patient with foot and ankle injuries sustained in combat. Foot Ankle Clin.
2010; 15: 175-186. 10.1016/j.fcl.2009.10.005
32. Owens, JG., Blair, JA., Patzkowski, JC., Blanck, RV. and Hsu, JR. Return to running and sports participation after limb salvage. J Trauma.
2011; 71: S120-S124. 10.1097/TA.0b013e3182219225
33. Patzkowski, JC., Blanck, RV., Owens, JG., Wilken, JM., Kirk, KL., Wenke, JC. and Hsu, JR. Comparative effect of orthosis design on functional performance. J Bone Joint Surg Am.
2012; 94: 507-515. 10.2106/JBJS.K.00254
34. Shawen, SB., Keeling, JJ., Branstetter, J., Kirk, KL. and Ficke, JR. The mangled foot and leg: salvage versus amputation. Foot Ankle Clin.
2010; 15: 63-75. 10.1016/j.fcl.2009.11.005
35. South BJ, Fey NP, Bosker G, Neptune RR. Manufacture of energy storage and return prosthetic feet using selective laser sintering. J Biomech Eng.
36. Sumiya, T., Suzuki, Y. and Kasahara, T. Stiffness control in posterior-type plastic ankle-foot orthoses: effect of ankle trimline.2. Orthosis characteristics and orthosis/patient matching. Prosthet Orthot Int.
1996; 20: 132-137.
37. Surve, I., Schwellnus, MP., Noakes, T. and Lombard, C. A Fivefold reduction in the incidence of recurrent ankle sprains in soccer players using the sport-stirrup orthosis. Am J Sport Med.
1994; 22: 601-606. 10.1177/036354659402200506
38. Vaughan, CL. and O’Malley, MJ. Froude and the contribution of naval architecture to our understanding of bipedal locomotion. Gait Posture.
2005; 21: 350-362. 10.1016/j.gaitpost.2004.01.011
39. Waters, RL. and Mulroy, S. The energy expenditure of normal and pathologic gait. Gait Posture.
1999; 9: 207-231. 10.1016/S0966-6362(99)00009-0
40. Wiley, JP. and Nigg, BM. The effect of an ankle orthosis on ankle range of motion and performance. J Orthop Sports Phys Ther.
1996; 23: 362-369. 10.2519/jospt.19184.108.40.2062
41. Wilken, JM., Rodriguez, KM., Brawner, M. and Darter, BJ. Reliability and minimal detectible change values for gait kinematics and kinetics in healthy adults. Gait Posture.
2012; 35: 301-307. 10.1016/j.gaitpost.2011.09.105
42. Wolf, SI., Alimusaj, M., Rettig, O. and Doderlein, L. Dynamic assist by carbon fiber spring AFOs for patients with myelomeningocele. Gait Posture.
2008; 28: 175-177. 10.1016/j.gaitpost.2007.11.012