A Contribution to Motorized Rehabilitation Devices
Alena Galajdova, Dusan Simsik, Boris Jobbagy and Marián Veselíny
Department of Automation, Control and Human-Machine Interactions, Technical University of Kosice,
Letná 9, Kosice, Slovakia
Keywords: Rehabilitation, Gait, Motorised Device, Chaotic Signal, Seniors.
Abstract: One of the most critical situations in life of seniors is a fall. Paper describes an experience of authors from
the development of a special rehabilitation tool for training of seniors who are in higher risk of falls.
Rehabilitation shoe SMILING is output of the international project supported by the European Commission.
Authors shortly describe the mechanical design of the motorised rehabilitation device controlled by a
chaotic signal. Paper is focusing on some results from the experimental work during clinical testing of the
prototype in Slovakia.
1 INTRODUCTION
Technological innovation, widely present in the field
of computer techniques, sensor networks, mobile
devices, contributes to the rehabilitation of persons
with mobility impairments. Every improvement in
the mobility is related to higher independence,
safety, in outside and home environment during
daily activities. European research in the field of
ICTs deals with the projects oriented to the
technologies creating intelligent environment and
embedded systems using wire and wireless
networks, different sensors of environment
parameters or user status in terms of health but also
identification of non-standard situations in daily
activities aim to increase the quality of life and
social inclusion of elderly as well as people with
disabilities.
There were several projects oriented on falling
and its prevention. The reason is that it is very
significant factor causing serious complications in
life of seniors, often ending by their death (Aizen
2007, Donald 2010, CDC 2011). We mention here
some of them showing variety of approaches applied
to the seniors fall prevention. Project Domeo was
focusing on communication with healthcare centre
through an assistive robotic system, and on physical
assistance for mobility functions using the
RubuWalker – intended for helping with walk,
sitting and rising, monitoring vital functions and
data processing (Canou, 2010).
Project VitaliShoe (shoe with embedded sensors)
developed a device that monitors movement during
the walk with focus on the prevention for falls and
injuries. Several types of sensors are installed in the
sole of the shoe to detect the obstacles and prevent
the fall among elderly.
Project BioSensing developed a sensor system
that quantifies simultaneously body acceleration,
knee angle, foot pressure, and repetitive loading
patterns of the knee joint during activities of daily
living. Patients get feedback if they move too much,
too little, or move in the wrong way. The sensor
system consists of a smart knee brace that measures
the knee angle, a combined angular velocity and
acceleration sensor, a foot sensor, a data acquisition
system, and a wireless communication system. Data
is uploaded to a webserver and presented via a web
application. Demonstrators were created for the
medical specialist (diagnostics), for the
physiotherapist (training), and patients at home
(training, monitoring), (Vlaskamp, 2011).
Another type of rehabilitation device was
developed in frame of the SMILING project. The
main goal was to develop a complex training system
for improvements of seniors’ stability using a new
mechatronic sophisticated rehabilitation shoes. Shoe
is a wearable rehabilitation device adjustable for
different feet sizes and different types of users from
the point of view of their abilities and mobility
functionalities. Such concept offers a flexible use of
the shoe for the wide variety of users.
323
Galajdova A., Simsik D., Jobbagy B. and Veselíny M..
A Contribution to Motorized Rehabilitation Devices.
DOI: 10.5220/0004935803230328
In Proceedings of the International Conference on Biomedical Electronics and Devices (TPDULL-2014), pages 323-328
ISBN: 978-989-758-013-0
Copyright
c
2014 SCITEPRESS (Science and Technology Publications, Lda.)
2 SMILING MOTORISED
REHABILITATION SHOES
The project SMILING main tasks were to develop
adjustable perturbation algorithms to fit individual
user’s specific needs and to implement a training
system to be used in rehabilitation, health care, and
fitness centres for a reorganization of the
rehabilitation process in ageing population.
The SMILING shoe is a complex mechatronic
system that requires interaction of various sensors
data, mechanical components, and human activity.
The SMILING shoe is worn on a standard shoe
used by user. The user has to react to changes of the
shoe inclinations to keep balance when walking
while completing specific tasks. Both left and right
shoes are equipped with 4 mechanical units driven
by DC-motors. Two are in the front and two are in
the back side. In generally, mechanisms change the
height after each or several steps, and in such way
they change inclinations of the shoes sole in two
planes – frontal and sagittal.
Two different mechanical designs of SMILING
shoes were developed: STRATH design (Figure 1)
(Carus 2010) and TUKE design (Figure 2), (Simsik,
2010).
Figure 1: STRATH design of the rehabilitation shoes.
Figure 2: TUKE design of the rehabilitation shoes.
Each rehabilitation shoe consists of a mechanical
unit with 4 drives, motor control unit MCU, swing
phase detector with wireless communication unit S-
module based on gyro and accelerometer and user
control unit UCU (Simsik, 2010).
The MCU must get and store, before each
training session, a suitable set of perturbations
patterns used for driving motors. It has to be
modified for every person reflecting his/her
functionalities and level of training. Driving of
motors by MCU must be synchronized with a human
walking activity that is detected by an external
accelerometer and gyroscope S-Sense processing
unit (Bulgheroni, 2009, Tacconi, 2010).
2.1 Chaotic Signal for Perturbations
of the Shoe Position
The motors change their position only during the
swing phase of gait cycle. Swing phase detection is
performed in real-time algorithm running in
electronics of the shoes, which processes signals
from internal sensors. These sensors, electronics and
wireless modules are embedded in IMEC modules
for wireless data transmission and S-module to
record 6-D gait parameters using gyroscopes and
accelerometers (Penders 2010).
The changes of perturbations are based on a
previous character of actions. Perturbations are
induced by the chaotic signal generating algorithms
and theory of dynamical systems (Figure 3).
Perturbations vary independently of the size of a
shoe to ± 10° in the sagital and frontal plane, and the
change in the sole height of up to 15 mm.
Figure 3: Generation of chaotic signal – speeded up.
2.2 Experiments with Shoes
A complex clinical testing of SMILING shoes was
provided in 4 countries with STRATH design.
0
10
20
30
40
50
-20
-10
0
10
20
-30
-20
-10
0
10
20
30
x-pert.
Lorenz Attractor
y-pert.
z-pert.
BIODEVICES2014-InternationalConferenceonBiomedicalElectronicsandDevices
324
The verification of SMILING rehabilitation
training influence on senior´s stability during
walking were published in several pilot studies in
terms of kinetic gait analysis, which confirmed the
positive impact of rehabilitation on the system
stability and dynamics of walking. (Marianni 2010,
Galajdova 2011). We describe below results from
standardised functional test provided during training
in Slovakia.
2.2.1 Evaluation the Effectiveness
of the System Operation
Physical and cognitive functional tests were
provided during trainings with SMILING shoes in 4
countries (Israel, Italy, Switzerland, and Slovakia).
For inclusion criteria were used:
Tinneti Performance Oriented Mobility
Assessment Tool (POMA), for risk of falls group
score below 26 (strictly) and for healthy control
group above 25 (strictly).
Mini Mental State Examination (MMSE), with
score ≥ 24
Any positive answer to the “Two-Questions case-
finding instrument for depression”, for healthy
controls group.
2.2.2 Standardized Functional Tests
Standardized functional tests were provided
following the recruitment stage and the performance
of the baseline tests (T0), a randomized controlled
cross-over trial was performed to assess the efficacy
of the SMILING system and training programme
(Figure 4).
Figure 4: Smiling training programme.
During the recruitment phase, 40 elderly were
screened, at the end of this phase, twenty elderly
individuals were included in the study. The subjects
who passed the enrolment process were 2 male and
18 female, with a mean age of 69,95±4,2 years, and
a mean weight of 70,38 ± 8,3kg.
The Tinetti assessment tool (POMA) is an easily
administered task-oriented test that measures an
older adult’s gait and balance abilities. The
participant is scored on 16 tasks (9 balance tasks and
7 gait tasks) graded on a scale where the maximum
score is 28. High score indicates higher levels of
balance function, while 19 or less indicates high risk
of fall. Across all Slovak participants at T0 the mean
score was 27,25 (STD = 0,8) and 100% of them
were included in the range of “low fall risk” (25-28).
This test shows no significant differences between
two groups after 4 weeks of training with Smiling
shoe – first group of 10 people and with Dummy
shoe (control group), second group of 10 people at
T1 and the same results we obtained also after
changing between these two groups of seniors at T2.
The 6-Minutes Walk test is a measure of
functional status (executed with S-Sense to obtain
gait analysis parameters) and was performed
indoors, along a flat, straight, 30m corridor, with a
hard surface. As regards the distance covered by the
subjects, all the participants finished the task in six
minutes without any stops, and the mean distance
covered at T0 is 426,53 m (STD = 61,0), from a
minimum of 272,5 m to a maximum of 521,6 m.
The Short Physical Performance Battery
(SPPB) assesses global physical performance and in
particular lower extremity function through balance
and gait tasks. The scoring system goes from a value
of 12 for the “best performance” to a value of 0 for
the “worst performance” and across all participants
the mean value is 11,00 (STD = 0,77) at baseline T0.
The Modified Narrow base walking test is an
easily administered task-oriented test that measures
an older adult’s gait stability and balance abilities
during both single and dual task conditions. Across
all 20 participants at baseline T0 the mean length of
stride was 0,61 m (STD = 0,06) in the single task
condition and 0,58 m (STD = 0,06) in the dual task
condition; the mean stride velocity was 0,867 m/s
(STD = 0,12) in the single task and 0,752 m/s (STD
= 0,16) in the dual task; regarding the mean step
error rate we can report that it was 0,023 (STD =
0,04) in the single task, 0,017 (STD = 0,02) in the
dual task. We noticed that during dual task the stride
velocity decreased.
The primary outcome of the study is distance,
measured during a self-paced 6 minutes walk and
length and velocity of stride during narrow base
walking test. The reason for this choice was the fact
that age-related decline in both gait speed and gait
stability are associated with increased fall risk in
older adults.
In Table 1 we can see parameters before first
training where participants are divided in two
groups, so we can compare parameters and see
consistency between these two groups.
AContributiontoMotorizedRehabilitationDevices
325
Table 1: Results for baseline T0 in Slovakia, Baseline
functional data (Mean ± STD; range) – Slovakia.
Mean ± STD; range
Smiling group Control group
Distance
covered in
6MWT (m)
426,36 ± 55,7 426,69 ± 71,8
SPPB 10,9 (STD =
0,88)
11,1 (STD =
0,74)
Narrow base
walking test
(length of
stride)
0,59 m (STD =
0,06
0,57 m (STD =
0,07
Narrow base
walking test
(stride velocity)
0,78 m/s (STD
= 0,19)
0,72 m/s (STD
= 0,14)
Table 2: Results for T1 in Slovakia, TI Functional data
(Mean ± STD; range) – Slovakia.
Mean ± STD; range
Smiling group Control group
Distance
covered in
6MWT (m)
426,90 ± 34,3 465,95 ± 93,09
SPPB 11,1 (STD =
0,74)
10,9 (STD =
0,74)
Narrow base
walking test
(length of
stride)
0,59 m (STD =
0,07
0,60 m (STD =
0,07
Narrow base
walking test
(stride velocity)
0,799 m/s (STD
= 0,18)
0,80 m/s (STD
= 0,15)
Table 3: Results for T2 in Slovakia, T2 Functional data
(Mean ± STD; range) – Slovakia.
Mean ± STD; range
Smiling group Control group
Distance
covered in
6MWT (m)
451,9 ± 62,12 478,6 ± 45,63
SPPB 11,1 (STD =
0,74
11,2 (STD =
0,63)
Narrow base
walking test
(length of
stride)
0,62 m (STD =
0,07
0,63 m (STD =
0,04
Narrow base
walking test
(stride velocity)
0,93 m/s (STD
= 0,22)
0,88 m/s (STD
= 0,13)
If we compare results for T0 mainly distance
covered during 6mwt and results from narrow based
walking test for stride velocity and for stride lenght
both groups were homogenous. At T1 after four
weeks of training with Smiling shoe (Smiling group)
and four weeks training of control group we can see
the differences: Control group achieved better result
in distance covered during 6mwt, and Smiling group
did not achieved better results. Also in narrow base
walking tests the results achieved by control group
were better. After T1 groups changed training
between Smiling and control group (Smiling group
performed training without Smiling shoes and
control group started use Smiling shoes, we
achieved results stated in Table 3. In T2 booth group
achieved better result in walking distance and also in
narrow base walking test (length of stride and stride
velocity).
As in Slovakia were tests performed only on
group of 20 seniors, the results did not show
statistically significant change. If we will take into
account all group from four countries (80
participants) we can see tendency still statistically
non-significant in stride length (Figure 5) and double
support (Figure 6) – two secondary outcome
parameters and important gait characteristics for gait
and balance quality.
Figure 5: Stride length.
Figure 6: Double support.
Figure 7 shows the stride velocity changes
throughout the trial. Clinically significant
improvement in the SMILING group over the first
BIODEVICES2014-InternationalConferenceonBiomedicalElectronicsandDevices
326
training period diminished, when analysed by the
mixed-model tests over the whole cross-over
protocol, revealing the absence of statistical
significance between the groups.
Figure 7: Stride velocity.
The assessment of functional status of subjects was
performed by standardized functional tests, such as
POMA, 6MWT, SPPB and Narrow Base Walking
Test.
The Tinneti Performance Oriented Mobility
Assessment Tool (POMA) aimed to assess the
mobility problems of the subjects. Higher scores of
the POMA test indicate better gait and balance.
Overall, mean POMA scores changed in both
training groups, but no difference achieved statistical
significance. Overall POMA balance score did not
increased significantly from 15,3 (at Smiling
Baseline) to 15,4 (at T2) and from 15,4 (at control
Baseline) to 15,6 (at T2). Overall POMA total score
changed from 27,1 (at Smiling Baseline) to 27,4 (at
T2) and from 27,4 (at control Baseline) to 27,6 (at
T2).
The Short Physical Performance Battery (SPPB),
based on standing balance, chair stand, and gait
speed, is a predictor of progressive disability,
hospitalization, nursing home admission and death
in the elderly population.
No significant changes were seen after
SMILING, nor in dummy trained subjects during the
whole experiment. The initial mean SPPB score was
quite high (the scale ranges from 0 to 12) and this
may be a reason for the absence of response of this
parameter to any training.
3 CONCLUSIONS
The rehabilitation shoes SMILING have an original
design based on sophisticated wearable individually
adjustable mechatronic unit controlled by a
microcontroller using the chaotic signal generated
by the Lorenz attractor. A complex of measurements
was provided during clinical testing in 4 countries.
Presented Slovak experience indicates that training
with SMILING shoes gives a positive impact on
improvement of users’ stability in walk.
Results of experiments showed some positive
effects on improving walking abilities of trained
persons. However, to prove the efficiency of the new
rehabilitation method will need more experiments,
and their evaluation including non-linear analysis.
ACKNOWLEDGEMENTS
The research leading to these results has received
funding from the European Community’s Seventh
Framework Programme (FP7/2007- 2013) under
grant agreement n°215493. The SMILING project
was coordinated by Dr.Fiorella Marcellini, INRCA
(Italian National Institute on Aging), Italy.
This work has been supported also by the Slovak
Grant Agency VEGA contract Nb. 1/0911/14
Implementation of wireless technologies into the
design of new products and services to protect
human health.
REFERENCES
Aizen E., Shugaev I., Lenger R., “Risk factors and
characteristics of falls during inpatient rehabilitation of
elderly patients.” Archives of gerontology and
geriatrics 2007;44(1):1-12.
Bulgheroni M, D'Amico E, Bar-Heim S, Carus D,
Harrison CS, Marcellini F. “The SMILING project:
Prevention of falls by a mechatronic training device.”
Proceedings of Telehealth and Assistive Technology
2009, Cambridge.
Canou, J.:, “Domestic robot for elderly assistance”,
http://ocg.at/AAL-F09/files/s1-zagler.pdf , 2010.
Carus, D., Hamilton, R., Harrison, C.,S.: Motorised Shoe
Mechanisms to Apply Chaotic Perturbations for Gait
Training. 2010. In ICABB-2010, Venice, (Italy).
CDC – Center for Disease Control and Prevention, “Falls
Among Older Adults”, An Overview 2011, (online),
2011, http://www.cdc.gov/homeandrecrea-
tionalsafety/falls/adultfalls.html
Donald,I.P., Christopher J. Bulpitt, “The prognosis of falls
in elderly people living at home”, http://
findarticles.com/p/articles/mi_m2459/is_2_28/ai_n275
49408/pg_4/
Galajdova A, Dolna Z, Simsik D. “A study of the
influence of mechatronic shoe design on gait pa-
rameters during stance phase.” Trendy v
biomedicínskom inžinierstve; 2011; pp 163-166.
AContributiontoMotorizedRehabilitationDevices
327
Mariani, B., Paraschiv-Ionescu, A., Aminian, K.: An
Instrumented 6 Minutes Walk Test: Assessment of 3D
gait variability for outcome evaluation in elderly
population. 2010. In Proceedings of the 1st
International Conference on Applied Bionics and
Biomechanics, Venice, Italy.
Mariani, B., Hoskovec, C., Rochat, C., Bula,C., and
Penders,J.: 3D gait assessment in young and elderly
subjects using foot-worn inertial sensors. 2010. In
Journal of Biomechanics, vol. 43, num. 15, p. 2999 -
3006.
Penders,J., van de Molengraft, J., Fabien Masse, Benoit
Mariani and Kamiar Aminian: S-sense: a wireless 6D
inertial measurement platform for ambulatory gait
monitoring. 2010. In: ICABB-2010, Venice, (Italy).
Šimšík D., Drutarovský M., Galajdová A., Galajda P.,
2010. Embedded Microcontroller Unit For Gait Reha-
bilitation Shoes. In: ICABB-2010, Venice, (Italy),
2010, pp.1-8.
Tacconi C., Paci G., Rocchi L., Farella E., Benini, L.,
Chiari L., “User Control Unit For The Smiling
System: Design And Functionalities”, In Icabb-2010,
Venice, (Italy), 2010, P. 2.
Vlaskamp, F., Cremers,G., Rianne Pas, Chris Arts, Bert
Bonroy, Ruud Gransier, Sytze Kalisvaart, Vincent
Kerkhofs, Peter Koomen, Rob van Lummel, Kenneth
Meijer, Hans Sauren, Mariëlle Swinkels, Bart
Vanrumste, Gerard Verschuren, Luc de Witte:
“BioSensing: using motion sensors in real life care
situations”, In: Assistive Technology Research Series –
vol. 29, AAATE-2011: p.48-56.
BIODEVICES2014-InternationalConferenceonBiomedicalElectronicsandDevices
328
