Preliminary kinematic evaluation of a new stance-controlknee–ankle–foot orthosis

Clinical Biomechanics 21 (2006) 1081–1089 www.elsevier.com/locate/clinbiomech

Preliminary kinematic evaluation of a new stance-control knee–ankle–foot orthosis Terris Yakimovich

a,b

, Edward D. Lemaire

b,c,*

, Jonathan Kofman

a,d

a

d

Department of Mechanical Engineering, University of Ottawa, Ottawa, ON, Canada b The Rehabilitation Centre (Ottawa Hospital), Ottawa, ON, Canada c Faculty of Medicine, University of Ottawa, Ottawa, ON, Canada Department of Systems Design Engineering, University of Waterloo, Waterloo, ON, Canada Received 24 February 2006; accepted 23 June 2006

Abstract Background. Stance-control knee–ankle–foot orthoses permit free knee motion in swing while providing knee flexion resistance in stance for individuals with quadriceps muscle weakness. However, some stance-control knee–ankle–foot orthoses require full knee extension to engage the knee-joint lock, thereby not providing knee support when climbing stairs or stepping over curbs. Stance-control knee– ankle–foot orthoses that do support a flexed knee are either heavy, bulky, expensive, offer a limited number of locking positions, or cause noise. This paper presents a preliminary kinematic evaluation of a new stance-control knee–ankle–foot orthosis that was designed to address these limitations.

Methods. Kinematic gait analysis was performed on three male knee–ankle–foot-orthosis users with knee extensor weakness in at least one limb (mean age: 56.3 years (SD 4.0)). Three walking trials were performed with the subjects’ current knee–ankle–foot-orthosis and then the new stance-control knee–ankle–foot orthosis (non-randomized before-after trial). Subjects completed a questionnaire about the new stance-control knee–ankle–foot orthosis and current knee–ankle–foot-orthosis.

Findings. A mean increase in knee flexion of 21.1° (SD 8.2) during swing and a greater total knee range of motion was found when walking with the new stance-control knee–ankle–foot orthosis. Two knee–ankle–foot-orthosis users experienced a reduction in pelvic obliquity and hip abduction angle abnormalities when walking with the stance-control knee–ankle–foot orthosis. Two out of three subjects preferred walking with the new stance-control knee–ankle–foot orthosis over their prescribed knee–ankle–foot-orthosis.

Interpretation. The new stance-control knee–ankle–foot orthosis permitted improved gait kinematics for knee–ankle–foot-orthosis users while providing knee support in stance and free knee motion in swing at appropriate instants in the gait cycle. Overall, the new stance-control knee–ankle–foot orthosis provided more natural gait kinematics for orthosis users compared to conventional knee– ankle–foot-orthoses. Ó 2006 Elsevier Ltd. All rights reserved.

Keywords: Gait; Orthotic devices; Knee–ankle–foot orthosis; KAFO; Stance control; Evaluation

1. Introduction Individuals with quadriceps muscle weakness due to polio, degenerative muscle disease, stroke, incomplete * Corresponding author. Address: Institute for Rehabilitation Research and Development, The Rehabilitation Centre, 505 Smyth Road, Ottawa, ON, Canada K1H 8M2.

E-mail address: elemaire@ottawahospital.on.ca (E.D. Lemaire).

0268-0033/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.clinbiomech.2006.06.008

spinal injury, congenital defects, or aging are often prescribed a knee–ankle–foot orthosis (KAFO) that locks the knee in constant full extension. Due to the absence of knee flexion, KAFO users must adopt abnormal gait patterns that can lead to premature exhaustion while walking (Waters et al., 1982), cosmetic problems, limited mobility, and pain and loss of motion at the hip and lower-back (McMillan et al., 2004). A new type of orthosis, referred to as a stance-control knee–ankle–foot orthosis (SCKAFO),

1082

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

has recently emerged to permit free knee motion in swing while providing knee-flexion resistance in stance.

Several gait studies have tested the efficacy of a variety of SCKAFOs. McMillan et al. analyzed the gait patterns and heart rates of three subjects with lower limb weakness walking with a SCKAFO built by Horton Technology Inc. (McMillan et al., 2004). The subjects participated in a series of gait analysis, obstacle course and treadmill trials. The study reported faster walking speeds, longer strides, less compensatory movements, increased mobility, and more symmetric gait patterns than walking with a fixed-knee KAFO. Hebert and Liggins (2005) studied one subject walking with a conventional KAFO and the Horton Technology Inc. SCKAFO. With the SCKAFO, the subject’s braced leg achieved near-normal knee flexion, reduced pelvic retraction and rotational excursion, and improved hip power generation. The subject’s contralateral limb experienced reduced abnormal ankle and hip power generation, increased knee power absorption, more typical quadriceps activation, and the elimination of vaulting.

Lehmann and Stonebridge (1978) investigated the effects of a bail-lock SCKAFO design on the oxygen consumption of two able-bodied subjects and two patients with spinalcord lesions. Both disabled subjects showed little improvement in energy expenditure since they did not have sufficient muscle strength to flex their knee adequately in swing or sufficient hip flexor strength to reach normal walking speeds.

Electromyographic analysis of 10 able-bodied subjects walking with the Intelligent Brace II measured a reduction in muscle activity when walking with the computer-controlled SCKAFO over walking with no brace at all (Tokuhara et al., 2000). The results suggested that the new brace could compensate for reduced lower limb muscle strength in walking.

The Mayo Clinic has performed numerous clinical evaluations on their wrap-spring SCKAFO. An examination of the gait kinematics and energy cost of a single poliomyelitis subject revealed an improvement in knee motion and metabolic energy requirements when wearing the SCKAFO compared to a conventional KAFO (Kaufman et al., 1996). In another study, a single able-bodied subject achieved higher cadence, walking velocity, and knee-flexion angle in swing when wearing the SCKAFO over walking with a fixed-knee KAFO (Irby et al., 1999). Irby et al. studied thirteen experienced KAFO users and eight novice KAFO users walking with the wrap-spring SCKAFO (Irby et al., 2005). The novice KAFO users demonstrated a significant increase in self-selected walking velocity and stride length walking with the SCKAFO over a fixed-knee KAFO. The experienced KAFO users typically walked at a reduced velocity and cadence with the SCKAFO over the conventional KAFO. The novice KAFO users generally achieved a significantly greater range of knee motion when walking with the SCKAFO than the experienced KAFO users. With the SCKAFO, both novice and experienced KAFO users experienced increased peak knee flexion in swing, and

reduced compensatory motions such as plantarflexion of the contralateral ankle in stance (vaulting) and dynamic pelvic obliquity (hip hike).

These studies have shown that SCKAFOs promote a more symmetric gait, improved mobility, improved gait kinematics, reduced compensatory movements, and/or reduced energy consumption for walking compared to conventional KAFO technology. Gait analysis was an effective tool for such SCKAFO evaluations. However, current SCKAFO knee-joints either required full knee extension to engage the knee-joint lock, created noise, offered a limited number of locking positions, or were considered by users to be heavy, bulky or expensive. Furthermore, some SCKAFOs limited where users could safely walk due to insufficient knee flexion resistance during activities that involved weight bearing on a flexed knee. Heavy SCKAFO designs led to excessive energy expenditure during ambulation and large dimensions offered poor cosmetic appeal.

Poor cosmesis is often a factor for decreased compliance.

Existing SCKAFOs also do not offer energy absorbing knee flexion in stance. Gait studies involving the Horton Technology SCKAFO and the wrap-spring SCKAFO found no natural stance phase-knee flexion on the braced side when walking. Elimination of stance-phase knee flexion can lead to abrupt loading of the limb, generating high forces on anatomical joints.

A new electromechanical SCKAFO (Yakimovich et al., 2006) (Fig. 1) was developed to address the functional, structural, and cost limitations of current commercial SCKAFOs. This orthotic knee-joint was designed to permit limited resisted knee flexion during initial limb loading, potentially providing energy-absorption to promote a smoother, more natural gait. No other SCKAFO approach provides resisted stance knee flexion. The new orthosis was designed to provide knee-flexion resistance during weightbearing while allowing knee extension at any knee angle.

The ability to extend the knee while weight-bearing is essential to securely climb stairs, traverse uneven ground, and recover from a stumble. The orthosis was designed to permit free-knee motion in flexion and extension during swing-phase. This new orthotic knee-joint is also 33% thinner medio-laterally and 7% lighter than Horton Technology Inc.’s Stance Control Knee Joint (McMillan et al., 2004), the most comparable joint structurally and functionally.

While this new SCKAFO has potential benefits over current technologies, initial gait analysis with KAFO users is necessary to verify that the new SCKAFO functions as designed. This study determined the effect of the new SCKAFO on KAFO-user gait and determined whether the new orthosis can correct KAFO-induced gait abnormalities.

2. Methods 2.1. SCKAFO prototype The newly developed SCKAFO uses medial and lateral knee-joints (Fig. 1) with a friction-based belt-clamping

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

1083

Fig. 1. Photograph of the newly designed SCKAFO: (a) showing the knee-joints installed on medial and lateral uprights, and (b) on a subject walking while wearing the new device.

mechanism that resists knee flexion when the belt is clamped (Yakimovich et al., 2006). When the foot begins to bear weight, pressure sensors located on the orthosis footplate detect pressure above a preset threshold. This deactivates a solenoid located at the knee-joint and permits belt clamping, thereby resisting knee flexion, throughout stance. Unresisted knee extension is permitted by the knee-joint mechanism at any time during stance. When the foot begins to leave the ground to initiate swing, the pressure detected by the sensors drops below the predetermined threshold.

This activates the solenoid to prevent belt clamping, and thus permits the knee-joint to flex and extend freely. For belt unclamping and free knee motion to occur, the SCKAFO user must apply a slight knee-extension moment to eliminate resultant knee-flexion moments.

Table 1 Subject data for the three KAFO-user subjects Subject 1

Subject 2

Subject 3

60 78 Polio

52 102 Post-polio

Walking aid

57 85 Polio, Parkinson’s Cane

None

Cane

Strength

Right

Left

Right

Left

Right

Left

Hip flexion Hip extension Knee flexion Knee extension Ankle dorsiflexion Ankle plantarflexion

4 5 5 5 5 5

4 4+ 3 3 4 3

4 3 4 3 4 4

5 5 5 5 5 5

5 5 5 5 5 5

4 4 5 3 4 5

Age (years) Weight (kg) Condition

Lower-limb muscle strengths are graded using the Oxford Scale (Coates and King, 1982).

2.2. Subjects Three male KAFO users of mean age 56.3 years (SD 4.0) and mass 88.3 kg (SD 12.3), participated in the study.

KAFO users with moderate knee extensor weakness in at least one limb were identified by the Prosthetics and Orthotics Service at The Rehabilitation Centre, Ottawa, Canada, and recruited by telephone by one of the researchers. Table 1 summarizes subject characteristics.

Subject 1 had previously been prescribed a conventional KAFO with a Becker LR-9002 Load Response knee-joint (Becker Orthopedic, Troy, Michigan, USA) that allowed up to 18° of spring-resisted knee flexion during limb loading. However, subject 1 was prone to trip, even on level ground, and he had insufficient strength to attain knee flex-

ion with the brace. His orthosis essentially functioned as a straight-leg KAFO. Therefore, subject 1 rarely wore his KAFO and instead preferred to walk with a walker or cane. Subject 2 had walked with a KAFO for the past 12 years. His KAFO allowed 18° of free knee flexion. Subject 2 almost always wore his KAFO when walking. Subject 3 was fitted with a Horton SCKAFO (Horton Orthotics Lab, Little Rock, AR, USA) five months prior to gait testing but found the free-motion triggering mechanism problematic and only used the brace when walking in busy environments or outdoors. For the purposes of this research, subject 3 switched his SCKAFO to the fully extended, locked position for the walking trials to provide similar function as a conventional KAFO.

1084

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

foot, lateral shank, lateral thigh, and posterior pelvis.

The new SCKAFO worn by a subject during clinical testing is shown in Fig. 1b. Three trials were captured of subjects walking along an 8-m walkway at a natural cadence, wearing their conventional KAFO. Subjects were then given approximately 20 min to practice walking with the SCKAFO. Threshold pressure levels for the three FSRs were adjusted at the beginning of the training period to accommodate the individual’s gait pattern and account for the constant residual pressures applied to the FSRs due to shoe tightness. To adjust FSR threshold levels, subject were asked to walk slowly and pause in terminal stance at the instant they wished the SCKAFO to switch to swing mode. A laboratory assistant then adjusted the individual FSR threshold pressure levels accordingly to achieve solenoid activation with this pressure profile. A similar process was used to determine the FSR pressure thresholds at the onset of limb loading for the stance phase. This method was effective in achieving a high level of control-system reliability in walking. Following the practice session, three SCKAFO trials for each subject were captured.

Video data were processed using the Ariel Performance Analysis System (Ariel Dynamics Inc., Trabuco Canyon, CA, USA) to obtain synchronized 3-dimensional (3D) coordinates of the tracking markers (Klein and DeHaven, 1995; Wilson et al., 1997). Synchronized 3D marker coordinates were exported as a C3D file to Visual3D (C-Motion Inc., Rockville, MD, USA). Visual3D was used to process the marker data for visual, graphical, and quantitative analysis. The mean stride time, stride length, walking speed, stance/swing duration ratio, and associated standard deviations were determined over the three trials.

Ensemble averages over three walking trials of the knee angle, sagittal hip angle, frontal hip angle, and pelvic obliquity were calculated for each subject, with the stride time normalized to 100% of the gait cycle. Using the pre-ensembled trial data, absolute peak kinematic values for each

2.3. Interventions Each of the three subjects was fitted with a custom SCKAFO by a certified orthotist and orthotic technician.

The same set of SCKAFO knee-joints, electronics, and foot switches was used in all three SCKAFOs. Three force-sensing-resistor (FSR) pressure sensors (Interlink Electronics, Camerillo, CA, USA) were adhered to the top surface of the SCKAFO footplate. The sensors were located at the heel, lateral forefoot and medial forefoot. A light emitting diode was connected in parallel to the solenoids to provide a visual signal when the solenoids were active. The light signal was recorded by a video camera and activation periods were determined during image processing. Footplate pressure sensors and knee-joint solenoids were remotely connected to a stationary control system via a 10 m cable.

Custom control software activated or deactivated the solenoids based on FSR pressure levels below or above a predetermined threshold.

Since the new SCKAFO offered selective knee motion, the orthotist included footplate and ankle joint structures on each custom SCKAFO which differed from the subject’s original KAFO. Tamarack articulating ankle joints (Tamarack Habilitation Technologies, Blaine, MN, USA) were used for subject 2 to allow resisted ankle rotation while constraining the ankle to a single plane of motion. Subject 1 had a 3/4 footplate added to his SCKAFO to help him create the knee extension moment needed to switch the SCKAFO to swing mode at the end of stance. Due to the limited range of motion for subject 3 s affected ankle, the orthotist used a fixed-ankle in the SCKAFO. A calibrated 4-camera, video-based motion analysis system was used to capture unilateral lower-limb kinematics. Passive reflective markers were placed on the test subjects using a modified National Institutes of Health (NIH) six degree-of-freedom marker set. Four markers were attached to segments on the braced side of each subject:

Table 2 Lower-body joint mean and standard deviation for peak angle (deg) and range of motion (deg) over three trials for the three subjects wearing their conventional KAFO and the new SCKAFO Subject 1 KAFO Peak angle (deg): mean (SD) K1 3.7 K2 1.1 K3 2.8 16.8 FH1 FH2 10.4 PO1 5.3 Mean range of motion (deg) Knee 4.3 Sagittal Hip 48.5 Frontal Hip 27.2 Pelvic Obliquity 9.4

(0.1) (1.0) (0.8) (1.9) (1.2) (0.2)

Subject 2 SCKAFO 0.0 2.7 14.5 17.0 4.1 4.6 17.5 36.3 21.0 8.5

(1.4) (1.5) (1.4) (0.7) (2.6) (1.2)

KAFO 1.9 0.8 19.6 18.8 10.8 5.1 20.3 39.3 29.7 8.2

(1.3) (0.1) (0.5) (0.6) (0.9) (0.3)

Subject 3 SCKAFO 3.8 2.8 44.3 9.5 16.9 3.3 47.0 37.7 26.4 4.2

(2.1) (0.2) (0.8) (0.7) (0.3) (2.0)

KAFO

SCKAFO

0.8 0.2 1.4 12.6 7.1 7.8

4.4 2.2 28.3 15.3 8.8 7.3

2.9 19.4 19.8 8.6

(0.4) (0.4) (0.8) (2.1) (1.8) (1.2)

(0.4) (1.6) (0.7) (1.1) (1.2) (0.2)

32.6 23.8 24.1 9.6

The peak value labels are defined as follows: K1, K2, K3: first, second and third knee angle peaks; FH1, FH2: first and second frontal hip angle peaks; PO1: pelvic obliquity peak. Peaks are shown in Figs. 2 and 3.

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

condition (KAFO, SCKAFO) were computed to yield mean values over the subject’s three trials. Mean joint ranges of motion and mean peak angles were compared between SCKAFO and KAFO trials for each subject.

These data were not ensembled over subjects since there was considerable variation in subject strength, walking ability, height, and age.

1085

3. Results 3.1. Kinematics The mean peak joint angles and mean ranges of motion over three trials for the three subjects wearing their conventional KAFO and the new SCKAFO are shown in Table 2.

Fig. 2. Mean knee angle over three walking trials with the prescribed KAFO and with the new SCKAFO for subjects (a) 1, (b) 2, and (c) 3. The gait cycle begins with initial floor contact of the braced limb. The shaded gray area indicates normal gait values ±1 standard deviation (Winter, 1991). KAFO SD denotes ±1 standard deviation from the KAFO condition. TO-SCKAFO denotes the mean instant of toe-off with the SCKAFO while TO-KAFO denotes the mean instant of toe-off with the KAFO.

1086

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

Graphs of mean knee angles over trials versus percent gait cycle for each of the three subjects are shown in Fig. 2.

Graphs of mean pelvic obliquity and frontal hip angles are shown in Fig. 3. Due to the different ankle joint and footplate structures used between the KAFOs and SCKAFOs, the kinematic ankle data was not considered in the gait analysis. Just before push-off, at approximately 45% gait cycle, peak knee flexion values (K2) were generally lower for the SCKAFO compared to walking with the KAFO, indicating that the leg was slightly more extended with the SCKAFO (Fig. 2, Table 2). The three subjects had considerably higher mean knee flexion (K3) during

swing when walking with the new SCKAFO compared to their prescribed KAFO. Peak K3 increased by 11.7°, 24.7°, and 26.9° for subjects 1, 2, and 3, respectively (Table 2) or by a mean of 21.1° over all subjects. This contributed to a mean knee range of motion increase of 13.2° for subject 1, 26.7° for subject 2, and 29.7° for subject 3 (Table 3), for an overall average increase in mean knee range of motion across subjects of 23.2° with the new SCKAFO. Subjects 1 and 2 experienced little knee stance–flexion with both their prescribed KAFO and the SCKAFO; however, subject 3 had an average increase of 5.6° for stance–flexion range of motion with the SCKAFO.

Fig. 3. Mean pelvic obliquity angle and frontal hip angle over three walking trials with the prescribed KAFO and with the new SCKAFO for subjects (a) 1, (b) 2, and (c) 3. The gait cycle begins with initial floor contact of the braced limb. The shaded gray area indicates normal gait values ±1 standard deviation (Cho, 2006). KAFO SD denotes ±1 standard deviation from the KAFO condition. TO-SCKAFO denotes the mean instant of toe-off with the SCKAFO while TO-KAFO denotes the mean instant of toe-off with the KAFO.

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

1087

Table 3 Mean and standard deviation for spatial and temporal stride parameters over three trials for the three subjects wearing their conventional KAFO and the new SCKAFO Subject 1 KAFO Stride parameters: mean (SD) Stride time (s) 1.73 (0.04) Stride length (m) 1.27 (0.02) Walking speed (m/s) 0.70 (0.03) Stance (% gait cycle) 58.3 (0.03)

Subject 2

Subject 3

SCKAFO

KAFO

SCKAFO

KAFO

SCKAFO

2.55 (0.07) 0.91 (0.03) 0.36 (0.00) 74.7 (0.22)

1.40 (0.04) 1.31 (0.05) 0.93 (0.06) 60.7 (0.06)

1.75 (0.03) 1.10 (0.05) 0.63 (0.02) 63.9 (0.14)

1.98 (0.22) 0.82 (0.07) 0.42 (0.08) 59.4 (0.10)

2.07 (0.18) 0.77 (0.11) 0.37 (0.09) 69.8 (0.08)

All three subjects experienced closer to normal frontal hip angles at toe-off (Fig. 3). Subject 1 experienced a 60.6% lower peak hip abduction in terminal swing, from 10.4° to 4.1° (FH2) (Table 2), when walking with the SCKAFO compared with his original KAFO. Subject 2 had less hip adduction when walking with the SCKAFO than with the KAFO, (9.3° less at the FH1 peak, Fig. 3b, Table 2). Walking with the KAFO, subjects 1 and 2 exhibited similar abnormal pelvic obliquity curves; both subjects consistently dropped their pelvis downward at 47% and 70% of the gait cycle (Fig. 3). The SCKAFO improved gait for subject 2 by eliminating the abnormal pelvic drop at 47% of the gait cycle while preserving the natural pelvic drop following toe-off at 70% gait cycle.

3.2. Spatiotemporal parameters Mean spatiotemporal stride parameters for the three subjects wearing their conventional KAFO and the new SCKAFO are summarized in Table 3. The mean walking speed over all subjects was 32% slower with the new SCKAFO compared to using the conventional KAFO. Walking with the SCKAFO resulted in a mean 17% shorter stride length, mainly for subjects 1 and 2, compared to walking with their prescribed KAFOs. KAFO users also had a mean 26% longer stride time and a mean 17% increase in stance percent of gait cycle when walking with the SCKAFO. The control system activated and deactivated the SCKAFO knee-joint locking mechanism at appropriate instants in the gait cycle. Since the knee naturally begins to flex before toe-off (Winter, 1991), the SCKAFO control system was set to allow free knee flexion at a specified foot pressure during terminal stance.

3.3. Participation feedback Following gait testing, a questionnaire on the user’s experience walking with the SCKAFO was given to each participant. Open-ended feedback by participants on how the SCKAFO compared to their existing brace was also obtained. Subjects 1 and 3 preferred the SCKAFO over their prescribed KAFOs as they felt the new orthosis required less effort to walk. Subject 2 preferred his prescribed KAFO to the SCKAFO because he subjectively reported an excessive amount of concentration to disen-

gage the SCKAFO knee-joint for free knee motion when initiating swing. The other two subjects found some increase in the amount of mental effort required when walking with the SCKAFO, specifically to disengage the SCKAFO knee-joint. Subject 1 expressed a desire for audio feedback when walking with the SCKAFO to assure him that the orthosis was switched to stance support mode.

Subject 3 found the SCKAFO lighter, less bulky, and easier to control than his prescribed Horton SCKAFO. All three subjects felt secure when walking with the SCKAFO. 4. Discussion The effectiveness of a new SCKAFO was evaluated using quantitative kinematic gait analysis and subjective participant feedback. The main functional objectives of the new SCKAFO were to provide free knee rotation in swing and provide knee stability in stance at any knee or ankle angle, while also allowing knee extension at any time.

This study demonstrated that the newly designed SCKAFO can achieve these functions. All three orthosis users achieved increased knee flexion when walking with the SCKAFO compared to their prescribed KAFO. The mean increase of 21.1° in K3 peak during swing over all subjects with the SCKAFO demonstrated an improvement in swing-phase gait patterns over KAFOs that limit knee flexion. Improvement with the SCKAFO was also shown by the increase in mean knee range of motion (ensemble mean increase: 23.2° or 488%). Past studies indicated that increased knee motion in swing leads to increased energyefficiency during walking (Kaufman et al., 1996; McMillan et al., 2004). The mean reported range of knee motion with a SCKAFO was 44° (Hebert and Liggins, 2005; Irby et al., 2005; Kaufman et al., 1996). While the new SCKAFO’s 32° average range of knee motion was lower than the range from the literature, this range may have been appropriate for the test subjects (i.e., the new SCKAFO did not inhibit natural knee motion at their self-selected walking speed).

Consideration must also be given to the other benefits provided by the new SCKAFO design; including, thinner and less bulky profile, lightweight structure, and resisted stancephase knee flexion.

By correcting the knee-flexion gait pattern, motion at the hip and pelvis changed, perhaps with slight improvement. Subject 1 experienced a decrease in hip abduction when walking with the SCKAFO, which indicated less

1088

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089

circumduction of his braced leg during swing. Subject 2 experienced an improved pelvic obliquity pattern. Hip and pelvis improvements were small and did not occur consistently among all subjects, making it difficult to ascertain whether the differences between conditions were anomalies or indeed benefits. Future studies involving more subjects may indicate whether the new SCKAFO improves hip and pelvic kinematics. McMillan et al. (2004) reported fewer extraneous pelvic movements and an increased smoothness of pelvic movement on both sides when walking with a SCKAFO. Irby et al. (2005) observed a reduction in maximum pelvic obliquity with their novice subjects and an overall reduction in pelvic obliquity range of motion when walking with a SCKAFO. While Irby et al.

(2005) did measure frontal plane hip motion, no improvement was reported with SCKAFO use.

The ability of the SCKAFO to provide resisted knee flexion upon limb loading allowed subject 3 to achieve smooth knee flexion during initial stance. This demonstrated that the SCKAFO can permit a more natural knee loading response than conventional fixed-knee KAFOs. This smooth resisted knee flexion allowed by the new SCKAFO in initial stance may benefit orthosis users who would otherwise find the rigid knee-locking in current SCKAFO designs uncomfortable, especially when load transmission from the ground to the thigh-pelvis-spine aggravates pre-existing soft tissue problems. Non-rigid knee flexion resistance may be especially beneficial in activities that apply relatively high loads to the limb, such as descending stairs or ramps. Future studies may focus on the efficacy and benefit of the SCKAFOs gradual knee flexion resistance in different activities.

Results from Subject 3 showed that the new SCKAFO will permit resisted knee flexion in stance. Subjects 1 and 2 did not achieve any notable knee stance–flexion, possibly because their gait had adapted to exclude stance-phase knee flexion when walking with their prescribed KAFO. Future studies may include special training for test subjects focussed on achieving stance-phase knee flexion.

All subjects found some initial difficulty to consistently provide the small extension moment required to disengage knee-joint locking; however, their experience walking with the SCKAFO increased over the short accommodation period in the laboratory and led to an increased ability to control the SCKAFO. The concentration required for SCKAFO control likely led to the overall shorter stride lengths, longer stride times, and therefore slower walking speeds compared to walking with their original KAFO. Because of the stationary nature of the SCKAFOs temporary control system, test subjects were only able to practice walking with the SCKAFO in the gait laboratory and were limited to approximately 20 min of SCKAFO training time. Ideally, the SCKAFO users should have several days or weeks to become accustomed to walking with the new SCKAFO to increase their ability and confidence to control the SCKAFO and improve gait performance.

Despite the short accommodation period and reductions in walking speed, two out of three KAFO users still preferred walking with the SCKAFO over their prescribed KAFO. It should be noted that nearly all commercial SCKAFOs require the user to provide a similar knee extension moment to initiate swing.

5. Conclusions The new belt-clamping SCKAFO design improved selected gait kinematics for KAFO users. The SCKAFO successfully provided knee support in stance and free knee motion in swing at appropriate instants in the gait cycle for all subjects. The SCKAFO also permitted a natural knee loading response, allowing natural knee flexion and knee extension in stance. While gait improved, additional training and accommodation time is required before SCKAFO users can walk without consciously extending their leg to disengage the locking mechanism. Two out of the three KAFO users preferred walking with the SCKAFO over their prescribed KAFO. Overall, the SCKAFO provided more natural gait kinematics compared to KAFO trials.

Future clinical studies should involve more subjects, bilateral gait analysis, and a longer accommodation period with the SCKAFO. Acknowledgments This research was funded by a University of Ottawa Interfaculty Collaborative Research Award and the Natural Sciences and Engineering Research Council of Canada (NSERC) – Idea to Innovation Program. The authors acknowledge the contributions of Rajiv Kalsi, Mark Russell, and Carla Serwotka for constructing the orthoses and offering their clinical expertise throughout the development and clinical testing stages of this research. We also acknowledge the valuable contributions of Shawn Millar, Joao Elias Tomas, Louis Goudreau, and Michael Munro, PhD, for their assistance and expertise.

References Cho, S., 2006. 3D gait analysis characteristics of the gait maturation process in normal Korean children. CGA Normative Gait Database.

Available from: <http://www.univie.ac.at/cga/data/index.html>, May 23.

Coates, H., King, A., 1982. The Patient Assessment: A Handbook for Therapists. Churchill Livingston, New York.

Hebert, J.S., Liggins, A.B., 2005. Gait evaluation of an automatic stancecontrol knee orthosis in a patient with postpoliomyelitis. Arch. Phys.

Med. Rehabil. 86, 1676–1680.

Irby, S.E., Kaufman, K.R., Wirta, R.W., Sutherland, D.H., 1999.

Optimization and application of a wrap-spring clutch to a dynamic knee–ankle–foot orthosis. IEEE Trans. Rehabil. Eng. 7, 130–134.

Irby, S.E., Bernhardt, K.A., Kaufman, K.R., 2005. Gait of stance control orthosis users: the dynamic knee brace system. Prosthet. Orthot. Int. 29 (3), 269–282.

Kaufman, K.R., Irby, S.E., Mathewson, J.W., Wirta, R.W., Sutherland, D.H., 1996. Energy-efficient knee–ankle foot orthosis: a case study. J. Prosthet. Orthot. 8, 79–85.

T. Yakimovich et al. / Clinical Biomechanics 21 (2006) 1081–1089 Klein, P.J., DeHaven, J.J., 1995. Accuracy of three-dimensional linear and angular estimates obtained with the Ariel Performance Analysis System. Arch. Phys. Med. Rehabil. 76, 183–189.

Lehmann, J.F., Stonebridge, J.B., 1978. Knee lock device for knee ankle orthoses for spinal cord injured patients: an evaluation. Arch. Phys.

Med. Rehab. 59, 207–211.

McMillan, G., Kendrick, K.K., Michael, J.W., Aronson, J., Horton, J.W., 2004. Preliminary evidence for effectiveness of a stance control orthosis. J. Prosthet. Orthot. 16, 6–13.

Tokuhara, Y., Kameyama, O., Kubota, T., Matsuura, M., Ogawa, R., 2000. Biomechanical study of gait using an intelligent brace. J. Orthop.

Sci. 5, 342–348.

1089

Waters, R.L., Campbell, J., Thomas, L., Hugos, L., Davis, P., 1982.

Energy cost of walking in lower extremity plaster casts. J. Bone Joint Surg. Am. 64, 896–899.

Wilson, D.J., Smith, B.K., Gibson, J.K., 1997. Accuracy of reconstructed angular estimates obtained with the Ariel Performance Analysis System. Phys. Ther. 77, 1741–1746.

Winter, D., 1991. The Biomechanics and Motor Control of Human Gait: Normal, Elderly and Pathological. University of Waterloo Press, Waterloo.

Yakimovich, T., Kofman, J., Lemaire, E.D., 2006. Design and evaluation of a stance-control knee–ankle–foot orthosis knee joint. IEEE Trans.

Neur. Syst. Rehabil. Eng. 14, 361–369.