http://informahealthcare.

com/idt
ISSN 1748-3107 print/ISSN 1748-3115 online

Disabil Rehabil Assist Technol, Early Online: 1–7

! 2014 Informa UK Ltd. DOI: 10.3109/17483107.2014.921938

RESEARCH PAPER

Attempt toward a development of aquatic exercise device

for gait disorders
Tasuku Miyoshi1, Fumie Komatsu1, Motoki Takagi1, and Noritaka Kawashima2
1
Faculty of Engineering, Iwate University, Iwate, Japan and 2Department of Rehabilitation for the Movement Functions, Research Institute of
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15

National Rehabilitation Center for Persons with Disabilities, Saitama, Japan

Abstract Keywords
Purpose: To develop an aquatic exercise device to facilitate locomotive motor output Aquatic gait training, elevation angle,
and achieve repetitive physiological gait patterns to improve movement dysfunctions. leg movement apparatus, locomotive
Methods: A custom designed leg movement apparatus (LMA) consisted of closed 4-linkage motor output, physiological gait pattern
mechanisms and one-length changeable link using a spring. Three-dimensional motions and
electromyographic (EMG) activities were recorded in eight healthy subjects to evaluate the History
reproducibility of the physiological gait patterns using the LMA with or without a spring
apparatus in water. Results: Using the LMA with a spring apparatus compared to walking in Received 24 September 2013
water, the foot trajectories and the time course of the elevation angles in each lower limb joint Revised 1 May 2014
kinematics were preserved. The time-series of the EMG showed reciprocal modulation between Accepted 4 May 2014
For personal use only.

agonist and antagonist muscle groups in the hip and ankle joints. However, the amplitudes of Published online 23 May 2014
the tibialis anterior muscle in the first half and rectus femoris in the last half of the movement
cycle were reduced using the LMA with a spring apparatus. Conclusion: We developed a novel
aquatic exercise device to reproduce physiological gait patterns. The LMA with a spring
apparatus would be particularly valuable in therapy for movement dysfunctions to facilitate
locomotive motor outputs.

ä Implications for Rehabilitation

 The leg movement apparatus with spring for underwater use (LMA) would be effective gait
training to induce the locomotor-like EMG activities.
 Hydrotherapy with the LMA has advantages over the partial body weight support treadmill
training on land with a robotic device; (1) the LMA is electrically and mechanically safe, and
(2) the LMA would require self-effort to generate the gait pattern for movement disorders, or
also enable passive gait training by the physiotherapists.

Introduction training devices that allowed the usage under water without an
exoskeleton device [3].
Hydrotherapy is one of the useful exercises for various orthopedic
Recent gait training methods for movement disorders to
disorders [1], since the impact force on the lower limb joints could
facilitate locomotive outputs were conducted on body weight
be easily controlled by varying the depth of immersion. However,
supported stepping movement on a treadmill [4–6] or with a gait
hip joint extension moment and hip extension electromyographic
orthosis [7,8]. For example, Wernig et al. [5] demonstrated that
(EMG) activities would increase against water resistance while
25 out of 33 incomplete spinal cord injured persons learned to
walking in water [2]. Therefore, it was necessary to monitor
walk independently at the end of 3–20 weeks (median 10.5).
walking speed and the immersion level during hydrotherapy,
According to the previous physiological studies, three somato-
especially in individuals with hip joint movement dysfunctions.
sensory feedback signals were essential for gait trainings: (1) the
To our knowledge, there was no evidence that hydrotherapy was
magnitudes of the EMG activities in lower limb muscles were
useful for individuals who could not walk independently, such as
related to the load applied to the foot [6], (2) sensory information
persons with complete spinal cord injury (ASIA-A). One explan-
of the load and hip flexion/extension position in generating and/or
ation for hydrotherapy not being used was that it was difficult for
shaping the rhythmic pattern [9] and (3) reciprocal leg move-
therapists to maintain patient safety while moving their legs under
ments, which mean the bipedal locomotion pattern in humans,
water. However, a more definitive reason was the lack of gait-
have a phase lag of 180 between the legs [10]. However, the
partial body weight supported treatment presents special
Address for correspondence: Assoc. Prof. Tasuku Miyoshi, Ph.D, Faculty challenges for the therapist. The passive movement of a disabled
of Engineering, Iwate University, Iwate, Japan. Tel: +81-19-6216357. person’s legs is ergonomically difficult since the individual being
Fax: +81-19-6216357. E-mail: tmiyoshi@iwate-u.ac.jp treated cannot support and/or move the legs independently.
 2 T. Miyoshi et al.
In order to improve this therapy, Colombo et al. [11] developed a
driven gait orthosis and the commercialized device called
Lokomat (Hokoma Inc., Volketswil, Switzerland), which can be
used on patients with varying degrees of paresis or spasticity for
up to one and a half hour. In addition, Hesse et al. [12] developed
the end-effector type of gait-training device to reproduce foot
trajectories as a pattern of normal gait motion. For further
powered training devices for gait impairments, see the review of
Marchal-Crespo and Reinkensmeyer [13]. It is noteworthy that all
powered exoskeleton type of gait training apparatuses have the
benefit of producing the physiological gait patterns but cannot
work under water without our previous orthosis [3].
We suppose that a combination of underwater gait training and
facilitating locomotive motor outputs is a very effective synthetic
training method. Therefore, a new gait-training device for use in
water that combines the benefits of traditional hydrotherapy and
the physiological gait training methods could be very effective for
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15
individuals who cannot walk independently. The purpose of this
study was to develop an aquatic exercise device to reproduce the
repetitive physiological gait patterns and to evaluate its efficacy
by measuring the lower limb angular displacements of the joint,
foot trajectories and lower limb surface EMG activities in healthy
subjects.
Methods
Concept and structure of the leg movement apparatus
The concept of the leg movement apparatus (LMA) is to
reproduce gait patterns that fulfill the following conditions:
(1) flexion/extension information of the hip joint from the hip
For personal use only.
mechano-receptors (for review, see [14]), (2) reciprocal relation-
ships between the right and left legs [10] and (3) load-related
somatosensory stimuli [6]. In this respect, the LMA should merge
with additional mechanisms to achieve kinematics, kinetics and
physiological requirements as follows: (1) the time course of
angular displacements of the joint as locomotion-like movements
in lower limbs for kinematics, (2) the modulation of the load
information onto the foot sole for kinetics and (3) the generation
of locomotor-like EMG activities as walking on land and in water.
The LMA consists of the newly developed closed 4-linkage
mechanisms as shown in Figure 1(A and B). It enables the
subjects to swing their legs by moving the handle while the handle
is connected to the footplate. Although the closed 4-linkage
mechanisms are not allowed to rotate around the metatarso-
phalangeal joint, the LMA enables the kicking form that appeared
at the late stance phase while walking, since the springs and
turnbuckles adjust the additional link length between the link and
the footplate. The length of the spring was adjusted by the
turnbuckle to generate the initial torque to elevate the footplate
around the hinge joint. The spring constant was set at
10.75 N/mm, and the initial length of the spring plus turnbuckle
is 412 mm. These sets produced 45 of rotation under 450-mm
backward slide movements without any loads.
Handling and evaluation in healthy subjects
Eight healthy subjects (27.5 ± 4.2 years, 172.5 ± 4.1 cm height,
64.6 ± 5.2 kg weight) participated in this study. They had no
history of orthopedic or neurological ailments and recent injury or
surgery that affected walking and/or standing upright. Each
subject provided informed consent acknowledging the experi-
mental procedures that were approved by the ethics committee of
the National Rehabilitation Center for Persons with Disabilities.
Before the trial sessions, all subjects were asked to practice the
LMA with a spring apparatus movement for 5 min so that the
subjects could experience the LMA movement. Using the LMA
Disabil Rehabil Assist Technol, Early Online: 1–7
with the spring apparatus, the subjects moved the handle 150 mm
forward and 100 mm backward from the initial position to
generate the reciprocal leg movement in a sinusoidal manner by
matching the movement frequency with the sound of a metronome
(0.6 Hz). The handle movement could induce approximately 20
and 15 of flexion and extension motions in the hip and 5 and
45 of dorsiflexion and plantar flexion motions in the ankle joints,
respectively. To examine the effect of the spring apparatus, a
session (control session) without a spring apparatus was also
conducted. The duration of each session was 1 min, and an
interval of over 1 min was taken for rest between each sessions.
Both sessions were conducted 10 times and the order of the trial
sessions was randomized. The water temperature was set to 34  C
and the water depth was set at 1.0 m (at the level of the pelvis, the
body weight was reduced approximately 50% by buoyancy).
A three-dimensional (3D) four-body segment model,
consisting of the pelvis, thigh, shank and foot, was defined
using custom-designed waterproofed active LED markers (6 mm
diameter, IS Engineering Co., Ltd., Ohdate, Japan) at the
following landmarks: iliac crest, greater trochanter, lateral
femoral condyle, lateral malleolus, lateral tarsal (heel) and fifth
metatarsal head. To check the arm, trunk and head movement, the
LED markers were also placed on the elbow, wrist, shoulder and
top of the head with a cap; these coordinates were not analyzed.
The LMA movement was measured with the same LED markers
to estimate the displacement and rotation of the footplate and the
length of the spring apparatus. All the markers were placed on the
right side of the leg and the LMA. The marker positions were
measured by a four-camera based system (OptiTrack FLEX:V100,
Natural Point Inc., Corvallis, OR) with a sampling frequency of
100 Hz and recorded. The cameras were placed on the right side
of the subject at an average distance of 2.0 m. Motion analysis
system software (VENUS3D, Nobby Tech. Ltd., Shibuya, Japan)
was used to reconstruct the marker positions into 3D coordinates.
A fourth-order zero-lag Butterworth filter was applied to reduce
noise in the coordinate data (cut-off frequency: 10 Hz).
Surface EMG activities were recorded from the muscle bellies
of the soleus (SOL), the medial gastrocnemius (MG), the tibialis
anterior (TA), the rectus femoris (RF) and the long head of the
biceps femoris (BF) muscles of the right leg using pairs of bipolar
electrodes (Ag-AgCl, 10 mm bar) with an inter-electrode distance
of 10 mm. These electrodes were waterproofed by wrapping
in microfilm. All EMG signals were amplified and band-
pass-filtered (from 30 Hz to 5 kHz) with a bioelectric amplifier
(AB-611J, Nihon Kohden Corporation, Shinjyuku, Japan). The
EMG signals were measured with a sampling frequency of 1 kHz
and stored on a personal computer via an AD converter (ML-880
Powerlab 16/30, AD Instruments, Colorado Springs, CO) for off-
line analysis. All EMG signals were full-wave-rectified after
subtracting the DC component for further analysis.
Data analysis
The body was modeled as an interconnected chain of rigid
segments and the angle of elevation [15] of the ith segment
(yi: thigh, shank and foot segment of the right leg) in the sagittal
plane with respect to the vertical plane was computed as:
 
xd  xp
i ¼ arctan
yp  yd
where p and d denoting proximal and distal end points of the
segment, and x and y denoting the horizontal and vertical
coordinates in the sagittal plane, respectively. Elevation angles
were positive in the hip flexion, knee extension and ankle
dorsiflexion direction relative to the vertical plane. Movement
cycle was defined as the duration time when the thigh elevation
 DOI: 10.3109/17483107.2014.921938 Underwater leg movement device for gait disorders 3
Figure 1. (A) A custom designed leg
movement apparatus (LMA) based on the
closed 4-linkage mechanisms used in this
study. Although the closed 4-linkage
mechanisms could not rotate around the
metatarsophalangeal joint, this apparatus
enables the kicking form at 50% of the
moving cycle using the spring. (B) Photo
of the newly developed LMA with a spring
apparatus. (C) The differences in the
movement pattern between attached or
unattached to the spring apparatus. From
30 to 70% of the movement cycle, the part
of the foot pedal moved upside and the knee
flexed using the LMA with a spring
apparatus. (D) The hip, knee, ankle, and foot
trajectories in the sagittal plane relative to the
instantaneous marker of the iliac crest
position using the LMA with or without
spring apparatus.
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15
For personal use only.

was observed as maximum hip flexion to the next time thigh
elevation was observed as maximum hip flexion. The angular
displacements of the joint, the changes in the length of the spring
apparatus and the EMG data without the first cycle in each session
were expressed in relation to the 100% cycle and then ensemble
averaged. The EMG amplitudes were normalized by the ensemble
averaged peak value of the control session. Student’s t test was
conducted to determine the significant differences under (1) peak
value of hip and knee flexion/extension angles and of ankle
plantar/dorsiflexion angles and the (2) mean EMG amplitudes of
first and last half in each muscle while using LMA with or
without spring apparatus. The levels were considered statistically
significant at p50.05.
We constructed the 3D position space plot of the movement
loops described by plotting the elevation angles of the hip,
knee and ankle versus the others, as already demonstrated [15].
To achieve the reproducibility of gait patterns using the LMA, the
grids correspond to the best-fitting plane, and its normal vector
was calculated, respectively, since the data points while walking
on land were projected close to the Eigen plane.
Results
Differences in the movement patterns with the LMA
Figure 1(C) shows the differences in the movement pattern using
the LMA with or without a spring apparatus. From 30 to 70% of
the movement cycle, the part of the foot pedal moved upside and
the knee flexed using the LMA with a spring device compared
with that of the trial session with no attached spring device.
Figure 1(D) shows stick pictures using the LMA without spring
device and with attached spring devices. It was clearly
demonstrated that the foot trajectories relative to the coordinate
of the marker of the iliac crest using the LMA with a spring
device was a normal gait form such as walking on land.
 4 T. Miyoshi et al. Disabil Rehabil Assist Technol, Early Online: 1–7
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15
For personal use only.

Figure 2. (A) Ensemble-averaged profiles of thigh (top), shank (middle) and foot elevation angle (bottom) throughout the one movement cycle in the
sagittal plane using the LMA with (thick line) or without spring apparatus (thin line) in subject #1. (B) The changes in the length of spring apparatus
with the LMA. The thin line denotes the pattern for subject #1 and the thick line demonstrates subject #2. (C) Ensemble-averaged profiles of EMG
activities throughout a movement cycle: soleus (top), medial gastrocnemius (middle upper), tibialis anterior (middle), biceps femoris (middle lower)
and rectus femoris (bottom) in subject #1.

Ensemble averaged profiles with the LMA

Figure 2(A) shows the effects of the attached spring device to the
angular displacements of the joint with the LMA. Using the
spring apparatus, the angular displacements of the ankle joint
show actual enhancement of the plantar flexion (15 ) at the
moment of the peak backward motion that was 50% of the
movement cycle. Also, the peak knee flexion increased simul-
taneously. Therefore, the first concept of our device to reproduce
the kicking form, such as walking, had been achieved using the
LMA with the spring apparatus. Unexpected angular profiles of
the joints were found in the hip joint, such as the decreased
extension throughout the movement cycle.
In the results, we could estimate the changes in the length of
the spring apparatus with the LMA for three out of eight subjects.
According to these results, the maximum changes in the length of
the spring apparatus appeared in the first half of the movement
 DOI: 10.3109/17483107.2014.921938 Underwater leg movement device for gait disorders 5
cycle (Figure 2B), which is the backward motion phase. Although data) using the LMA with and/or without the spring apparatus.
the amplitude of the length changes depended on each subject, the The main difference in the Eigen plane was due to the increased
modulation pattern was preserved; peak changes appeared in the range of motion of the ankle, simultaneous changes in the knee
first half of the movement cycle, and the length changes were and hip motion from 30 to 70% of the movement cycle using the
not constant throughout the movement cycle. In this respect, the
spring apparatus actually made ankle plantar flexion, since
the timing of the peak amplitude of the length changes in the
spring apparatus was the initiation of the ankle plantar flexion
phase.
Figure 2(C) shows typical ensemble averaged profiles of each
EMG activity with or without the spring apparatus, respectively.
With the first half of the movement cycle, the SOL, MG, BF and
RF activities were preserved using the LMA with a spring
apparatus. In the case of the TA, the activity levels were
significantly increased using the LMA with the spring apparatus.
In addition, the peak SOL, MG and BF activities appeared at the
moment of peak plantar flexion (50% of the movement cycle).
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15

Conversely, the SOL and RF EMG activities decreased in the last

half of the movement cycle using the LMA with a spring
apparatus. In the same phase, significant differences in the
modulation pattern were not found in MG, TA and BF EMG
activities, respectively. The reciprocal EMG modulation patterns
throughout the movement cycle appeared between agonist and
antagonist muscle groups with the LMA: ankle plantarflexion
(SOL and MG) versus ankle dorsiflexion (TA) muscle and hip
extension (BF) versus hip flexion (RF) muscle. All of these EMG
patterns were found in all subjects. Table 1 shows the summarized
differences in kinematic and EMG using the LMA with or without
spring apparatus.
For personal use only.

Reproducibility of the physiological gait pattern

It was observed that the relative relationships between each
elevation angle while walking on land converged into an Eigen
plane in 3D space [15]. Figure 3 represents the relative
relationships in 3D plots of angular displacements of the joint
and made best-fit planes while walking on land (unpublished

Table 1. Differences in kinematics and EMG activities while LMA with

or without spring apparatus.

With Without Significance

Hip [deg]
Flex. 25.0 ± 1.28 28.06 ± 2.91 n.s.
Ext. 6.2 ± 1.45 6.25 ± 1.99 **
Knee [deg]
Flex. 29.9 ± 2.05 39.74 ± 4.23 **
Ext. 2.51 ± 1.76 9.04 ± 2.37 *
Ankle [deg]
Plantar 62.53 ± 2.19 47.95 ± 4.21 **
Dorsi 70.22 ± 3.89 72.33 ± 3.55 n.s.
SOL (ratio)
First half 0.21 ± 0.05 0.26 ± 0.04 n.s.
Last half 0.28 ± 0.07 0.44 ± 0.11 *
MG (ratio)
First half 0.47 ± 0.14 0.42 ± 0.18 n.s.
Last half 0.32 ± 0.07 0.31 ± 0.09 n.s.
TA (ratio)
First half 0.69 ± 0.22 0.27 ± 0.06 *
Last half 0.21 ± 0.03 0.28 ± 0.12 n.s.
BF (ratio)
First half 0.45 ± 0.11 0.46 ± 0.09 n.s.
Last half 0.39 ± 0.13 0.41 ± 0.11 n.s.
RF (ratio)
First half 0.22 ± 0.08 0.21 ± 0.07 n.s.
Last half 0.32 ± 0.06 0.52 ± 0.08 **
Figure 3. Three-dimensional plots of angular displacements of the joint
n.s.: no significance. and best-fit planes while walking on land (top), using the LMA without
*p50.05, **p50.01. spring apparatus (middle) and with a spring apparatus (bottom).
 6 T. Miyoshi et al. Disabil Rehabil Assist Technol, Early Online: 1–7

LMA with the spring apparatus. It was clearly demonstrated that
the spring apparatus reproduced the normal vector of the Eigen
plane compared with that of land walking: hip, knee and ankle
were positive, negative and 1.0 for walking on land; negative,
positive and 1.0 without the spring apparatus; and positive,
negative and 1.0 with a spring apparatus, respectively.
Discussion
The main objectives of this study were to develop a novel
aquatic exercise device to reproduce physiological gait
patterns and to investigate the lower limb kinematics, kinetics
and EMG activities while using this device. The results were as
follows: (1) the foot trajectories and the time course of the
elevation angles in each lower limb joint kinematics using the
LMA with the spring device were preserved such as those of
walking on land; (2) the phasic modulation of the length of the
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15
patterns of the MG and BF muscles did not depend on the spring
apparatus.
The increased RF EMG activities in the last half of the
movement cycle were expected, since the RF EMG activities
increased while backward walking [15]. In the water, the
amount of the water resistance depends on the size of the
surface area of the moving object and needs to propel the
leg in the forward direction against water resistance. Conversely,
it was clearly demonstrated that the LMA with the spring
apparatus could reduce the RF EMG activities, especially in
the 60–80% movement cycle phase. These obtained effects
were expected, since the enhanced knee flexion using the
LMA with a spring apparatus, the area of the lower limb segment
in the frontal plane should reduce compared with those without
a spring.
Using the LMA with a spring apparatus, the increased
EMG activities of the TA EMG activities in the first half of

spring apparatus could be achieved; (3) the time-series of the

the movement cycle and the SOL in the last half were
EMG showed reciprocal modulation between agonist and antag-
unanticipated. Further research was necessary; however, it
onist muscle groups in the hip and ankle joints; and (4) the
would reflect maintenance of the ankle angular displacement or
amplitudes of the TA in the first half and RF in the last half of the
foot position on the footplate, since the LMA lacks a foot fixation
movement cycle were reduced using the LMA with a spring
device.
apparatus. It was noteworthy that the LMA could be used under
severe experimental conditions, even while submerged, without
Implication of gait rehabilitation
disturbing gait patterns. However, the extent to which the findings
from this study can be applied to patients with movement Recent studies demonstrated that to induce locomotor-like EMG
dysfunctions is unclear because the data are based on experiments activities in lower limb muscles is essential to regaining neural
limited to healthy subjects. coordination of gait rehabilitation for spinal cord injured patients
[18–20]. In addition, enhancing the locomotor-like EMG
Kinetic behaviors using the LMA activities depended on the phasic modulation [10] and the
For personal use only.

amount of the load information onto the foot sole [21].

Ivanenko et al. [16] demonstrated that the foot trajectories of body
Although there were differences in the modulation pattern of
weight support treadmill walking were preserved, whereas the
the load information, which was depicted by the changes in the
patterns changed significantly only when the load information
length of the spring apparatus of the LMA as shown in
from the foot was unavailable. In this respect, the preserved foot
Figure 2(B), using the LMA would be effective gait training to
trajectories and the time course of the elevation angles were
induce the locomotor-like EMG activities.
considered the phasic modulation of the length of the spring
The value of partial body weight support treadmill training
apparatus with the LMA. The results showing the differences in
(BWSTT) for disabled individuals has been well established,
the normal vector of the 3D plot, however, reflect the differences
especially when initiated soon after an injury. However, recent
in the profiles of the angular displacement of the joint with the
studies demonstrated that stroke patients who receive robotic and
LMA, especially in the ankle plantar flexion around 50% of the
passive training in combination with physiological and voluntarily
movement cycle. Reconstructing the physiological gait patterns,
therapies achieve better motor function than training without or
such as walking on land, should reproduce the ankle plantar
only with a robotic device [22]. In this respect, hydrotherapy with
flexion at the kicking phase (backward motions using the LMA
the LMA has at least three advantages over the partial BWSTT on
with the spring apparatus). Although further improvements should
land with a robotic device. First, the LMA is electrically
be needed, aquatic exercise using the LMA would be effective for
and mechanically safe since it consists of closed four-linkage
gait rehabilitation as well as for general gait-training for patients
mechanisms with no electrical actuators. Second, the body weight
with movement disorders.
is managed by changing the depth of immersion, which allows
the patient to maintain an upright posture. And finally, the
The time-series of the phasic EMG modulation pattern LMA would require self-effort to generate the gait pattern
using the LMA for movement disorders or also enable passive gait training by
the physiotherapists. Further studies should determine whether or
The reciprocal modulation between agonist and antagonist lower
not the LMA is beneficial to regaining the physiological gait
limb muscle groups while walking on land is obvious. Thus, the
pattern for gait impairments, such as spinal cord injured and/or
preserved time series of the phasic EMG activities between
stroke patients.
triceps surae and TA or between BF and RF using the LMA
demonstrates the neurologically natural gait phenomena. In the
case while walking in water, however, Miyoshi et al. [17] Conclusion
demonstrated that dissociative activation patterns occurred in the We developed a novel aquatic exercise device to reproduce the
triceps surae muscle of subjects walking in water; the MG EMG physiological gait patterns and evaluated its efficacy with
activities were enhanced with an increase in walking speed. It was kinetics, kinematics, and EMG data. The LMA would be
also well observed that both the MG and BF EMG activities particularly valuable in therapy for movement dysfunctions to
increased with the increase in movement speed while walking in facilitate locomotive motor outputs.
water [2]. This results that the modulation patterns of the MG and
BF EMG activities were preserved while using the LMA with
Declaration of interest
and/or without the spring apparatus under constant movement
frequency. It was considered that the preserved modulation None of the authors have conflicts of interest.
 DOI: 10.3109/17483107.2014.921938
References
1. Eitner D. Exercise in water. In: Kuprian E, ed. Physical therapy for
sports. Philadelphia: Saunders; 1982:155–7.
2. Miyoshi T, Shirota T, Yamamoto S-I, et al. Effects of the walking
speed to the lower limb joint angular displacements, joint moments
and ground reaction forces during walking in water. Disabil Rehabil
2004;26:724–32.
3. Miyoshi T, Hiramatsu K, Yamamoto S-I, et al. Robotic gait trainer in
water: development of an underwater gait-training orthosis. Disabil
Rehabil 2008;30:81–7.
4. Wernig A, Müller S. Laufband locomotion with body weight support
improved walking in persons with severe spinal cord injuries.
Paraplegia 1992;30:229–38.
5. Wernig A, Müller S, Nanassy A, Cagol E. Laufband therapy based
on ‘‘rules of spinal locomotion’’ is effective in spinal cord injured
persons. Eur J Neurosci 1995;7:823–9.
6. Harkema SJ, Hurley SL, Patel UK, et al. Human lumbosacral spinal
cord interprets loading during stepping. J Neurophysiol 1997;77:
797–811.
Disabil Rehabil Assist Technol Downloaded from informahealthcare.com by Kainan University on 04/04/15
7. Kojima N, Nakazawa K, Yano H. Effects of limb loading on the
lower-limb electromyographic activity during orthotic locomotion in
a paraplegic patient. Neurosci Lett 1999;274:211–13.
8. Nakazawa K, Kakihana W, Kawashima N, et al. Induction of
locomotor-like EMG activity in paraplegic persons by orthotic gait
training. Exp Brain Res 2004;157:117–23.
9. Dietz V, Müller R, Colombo G. Locomotor activity in spinal man:
significance of afferent input from joint and load receptors. Brain
2002;125:2626–34.
10. Kawashima N, Nozaki D, Abe MO, et al. Alternate leg movement
amplifies locomotor-like muscle activity in spinal cord injured
persons. J Neurophysiol 2005;93:777–85.
11. Colombo G, Joerg M, Schreier R, Dietz V. Treadmill training of
paraplegic patients using a robotic orthosis. J Rehabil Res Dev 2000;
For personal use only.
37:693–700.
Underwater leg movement device for gait disorders 7
12. Hesse S, Uhlenbrock D, Werner C, Bardeleben A. A mechanized
gait trainer for restoring gait in nonambulatory subjects. Arch Phys
Med Rehabil 2000;81:1158–61.
13. Marchal-Crespo L, Reinkensmeyer DJ. Review of con-
trol strategies for robotic movement training after
neurologic injury. J Neuroeng Rehabil 2009;6:20 (open access).
14. Dietz V. Proprioception and locomotor disorders. Nat Rev Neurosci
2002;3:781–90.
15. Grasso R, Bianchi L, Lacquaniti F. Motor patterns for human gait:
backward versus forward locomotion. J Neurophysiol 1998;80:
1868–85.
16. Ivanenko YP, Grasso R, Macellari V, Lacquaniti F. Two-thirds
power law in human locomotion: role of ground contact forces.
Neuroreport 2002;13:1171–4.
17. Miyoshi T, Sato T, Tanizaki M, et al. Altered activation pattern
in synergistic ankle extensor muscles in the reduced-gravity
environment. Gait Posture 2006;24:94–9.
18. Nakazawa K, Kakihana W, Kawashima N, et al.
Induction of locomotor-like EMG activity in paraplegic
persons by orthotic gait training. Exp Brain Res 2004;157:
117–23.
19. Kawashima N, Nozaki D, Abe MO, Nakazawa K. Shaping
appropriate locomotive motor output through interlimb neural
pathway within spinal cord in humans. J Neurophysiol 2008;99:
2946–55.
20. Kawashima N, Nakazawa K, Akai M. Characteristics of the
locomotor-like muscle activity during orthotic gait in paraplegic
persons. Neurol Res 2008;30:36–45.
21. Dietz V, Harkema SJ. Locomotor activity in spinal cord-injured
persons. J Appl Physiol 2004;96:1954–60.
22. Waldner A, Tomelleri C, Hesse S. Transfer of scientific concepts to
clinical practice: recent robot-assisted training studies. Funct Neurol
2009;24:173–7.