Investigation of the Sensorimotor Training using Wireless Sensor
Networks
Analyzing Three Different Exerciser
A. Thiers
1
, L. Meffofok
1
, K. Orlowski
1
, K. Schrader
2
, B. Titze
3
, A. l’Orteye
3
and T. Schrader
1
1
Brandenburg University of Applied Sciences, Department of Informatics and Media, Brandenburg, Germany
2
SAfP Spektrum Akademie f
¨
ur Physiotherapie, Berlin, Germany
3
St
¨
adtisches Klinikum Brandenburg GmbH Akademisches Lehrkrankenhaus der Charit
´
e,
Abteilung Medizinische Schule, Brandenburg, Germany
Keywords:
Ankle Injuries, Sensorimotor Training, Proprioception, EMG, ECG.
Abstract:
The sensorimotor training method is more and more applied in therapy, rehabilitation, prevention as well as
to increase performance. The training comprises of the practice of motor action and the improvement of
the proprioception. The increasing demand for training has led to a growing range of training equipment.
However, this can only be achieved when the training equipment is applied correctly. The application of the
training equipment is influenced by two factors: the indiviual behavior when using the equipment as well as
the subjective impression of effort by each patient, both of which can differ strongly. For the purpose of this
study, 18 healthy test persons were recruited. The test comprised of each person having to use each device
at various levels of difficulty. Physiological data measured by EMG and ECG as well as psychological data
were collected. The study revealed that the equipment demands different levels of effort from the individual
depending on the physical abilities. Furthermore, it was shown that changing tasks on the particular exerciser
improved the effort of the test persons.
1 INTRODUCTION
Typical fields of sensorimotor training are the pre-
vention of ankle sprains, the therapy of instability
of joints, the rehabilitation of cruciate ligament in-
juries and the improvement of the athletic perfor-
mance (H
¨
afelinger and Schuba, 2010).
Reflexes, rhythmic and cyclical motion patterns as
well as controlled voluntary movements can be sum-
marized under the term motor activity. Besides the
motor activity, the so-called proprioception, depth
sensitivity, are the main characteristics of a sensori-
motor training (Lukas et al., 2011).
The training’s objectives include the regeneration
and the activation of injured proprioceptors, the in-
crease of their potential force and the reduction of the
response time of the stabilizing musculature. Hence,
the overall goal of the training is the improvement of
the reception and the processing of information. In
addition, the translation of the information into the
targeted movement is necessary to guarantee the op-
timal execution of the movement. The positive health
benefit of this training method has been verified in
several scientific studies (Dohm-Acker et al., 2008;
Lukas et al., 2011). Therefore, the market for senso-
rimotor training equipment is still growing.
To maximize training results a correct application
of the equipment is paramount. Due to the different
materials, functions and characteristics of the exer-
cisers correct application might be challenging. An
additional difficulty for the therapy planning is repre-
sented in the subjective impressions by the patients.
When facing the same task, different opinions regard-
ing the strain of the exercise tend to be expressed by
the patients. In the following experimental study the
mode of action of three different exercisers was ana-
lyzed. The research involved 18 subjects of different
age and with different experiences.
2 MATERIAL & METHODS
2.1 Measurements
The Shimmer
TM
measurements are wireless sensors
413
Thiers A., Meffofok L., Orlowski K., Schrader K., Titze B., l’Orteye A. and Schrader T..
Investigation of the Sensorimotor Training using Wireless Sensor Networks - Analyzing Three Different Exerciser.
DOI: 10.5220/0004329004130416
In Proceedings of the International Conference on Health Informatics (HEALTHINF-2013), pages 413-416
ISBN: 978-989-8565-37-2
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
with a small form factor. Using Bluetooth allows
for online data streaming in real-time. Providing
pre-amplification of EMG-Signal the non-invasive
method represents the whole activity of a muscle us-
ing two or three channel data acquisition. Similar
to the EMG signal, the ECG measures the electri-
cal activity of the heart muscles with a two channel
ECG. As the EMG data, the ECG signal is also pre-
amplified (Shimmer Research Support, 2012b; Shim-
mer Research Support, 2012a).
2.2 Exercisers
To evaluate different sensorimotor training methods
three different exercisers were chosen and evaluated
under different aspects (fig. 1).
Figure 1: Balance Pad, Balance Board, Ortho Pad.
The Balance Pad offers multifaceted applications
of the sensorimotor training. The Pad is made of a
special foam. One characteristic is its damping prop-
erty. Hence, the destabilization of the lower extremi-
ties is supported (Sport-Thieme, 2012a).
The top of the Balance Board is made of stable
and reinforced plastic with a diameter of 40 cm. Here
the aim is to strengthen the musculature of buttocks,
legs, back and abdomen (Sport-Thieme, 2012c).
The Ortho Pad is a specific type of a teeter board. Its
special construction offers a multi-axial balance train-
ing, as the patient has to handle a balance shift in at
least four directions (Sport-Thieme, 2012b).
2.3 Experimental Setup
18 healthy subjects from the university and the medi-
cal school took part in the study. They gave their writ-
ten consent to participate after being informed about
the test procedure. Based on age and skills, a divi-
sion into three groups was made. One group consisted
of young students with no experiences in sensorimo-
tor training. Group two comprised of students with
comprehensive experience in balance training. Group
three included all subjects over the age of 40 years
irrespective of their experience with the exercisers.
During the research, two different types of
Shimmer
TM
sensors were used. For stress documen-
tation the ECG sensor was placed on the anterior tho-
rax for measuring the chest leads. During the data
analysis, primary attention was paid to the muscular
activity. The ECG was of secondary importance. To
examine the muscular activity, sensors were placed
on the right and on the left lower extremities and but-
tocks. The placement of the sensors was chosen due
to the fact that injuries of the ankle joint are often
treated with proprioceptive training and it is therefore
of interest to look at the behavior of the M. tibialis an-
terior. This muscle supports dorsal flexion, supination
and adduction of the foot (SENIAM project, 2012).
To proof the assumption that the upper body areas par-
ticipate in balancing motions, the EMGs of the right
and the left M. gluteus maximus were also measured.
All subjects had to perform the same test pro-
cedure for all exercisers. The order of the training
equipment was randomized for each test person. One
test sequence comprised of a reference measurement
(30 s) and the measurement while on the test equip-
ment (210 s). The reference measurement was made
on the floor while the subject was in rest. Immedi-
ately afterwards, the measurement on the exerciser
took place. The test scenario consisted of five phases
of varying difficulty. Each of the first two phases
(standing on the exerciser with eyes open and eyes
closed) lasted 30 s. The following phases, throwing
a medicine ball and doing a cable pullover, had a du-
ration of 60 s each. The final phase is identical to
the first one regarding the task as well as the duration.
After finishing the procedure on one exerciser the test
persons were asked different questions regarding the
handling of the equipment in general and which phase
was the most exhausting one. All questions used an
eleven-point answer scale (0 equals “no effort” to 10
equals “very high effort”) except for the one inquiring
about the most exhausting phase.
2.4 Data Analysis
The data analysis process startet with the normaliza-
tion of the EMG data. The absolute values were trans-
formed into relative values by using the data of the
reference measurement. A notch filter with a block-
ing frequency of 50 Hz and a band-pass filter from 15
to 500 Hz were applied to the raw EMG (Merletti and
Parker, 2004). In the next step, EMG signal process-
ing required the full-wave rectification of the signal.
In order to analyze the data in the time domain, differ-
ent statistical parameters, such as mean and maximum
were calculated. On one hand, the parameters were
computed over an interval of three seconds. On the
other hand, the parameters in each of the phases were
calculated. Similar to the processing of the EMG sig-
nal, the ECG was also bandpass (0.05 Hz - 30 Hz )
and notch filtered (Husar, 2010). The properties of
the notch filter were identical to those of the EMG.
Analyzing the ECG in time domain includes the cal-
culation of the heart rate itself as well as the compu-
HEALTHINF2013-InternationalConferenceonHealthInformatics
414
tation of statistical parameters of the heart rate.
3 RESULTS
The aim of the sensorimotor training is to improve
the motor action and the proprioception. The train-
ing process should lead to a targeted use of the mus-
cles. Therefore, it is to be expected that trained sub-
jects have both, a smaller total amount and a shorter
time in muscle activity. This effect is visualized in
figure 2, comparing one untrained subject from group
one with one trained subject from group two. Already
in the phase “Eyes Open” the difference between the
trained and untrained subjects is clearly illustrated.
The trained person has lower voltage values across
all phases. In addition it is visible that the subject of
group two restored its balance faster and was able to
keep the balance much better because there was tem-
porarily nearly no muscle activity. Especially in the
phases “Eyes Open” and ”Cable Pullover” the com-
parison of the individual EMG courses show a clear
difference in the patterns. For example, in the phase
“Cable Pullover”, the untrained subject has to apply
more than double the muscle activity. A similar effect
can be seen in the first ans last phase. In summary,
figure 2 supports the assumption that trained persons
can better preserve their balance and are also able to
restore their equilibrium faster.
Figure 2: Comparison of one representative of group one
and two.
Figure 3 documents the muscle activity of an un-
trained young subject on the various equipments. It is
obvious that the three exercisers lead to different mus-
cle activity patterns. Comparing the same phases with
the three pieces of equipment revealed that each exer-
ciser required different skills of the subject. Both, the
Balance Board and the Ortho Pad demanded a high ef-
fort during the “Eyes Closed” phase. On the contrary,
the comparison of the sum of the voltage values of the
Balance Pad in this phase showed relatively low val-
ues. The third and the fourth phase illustrated regular
muscle activity. At the start of both phases the volt-
age values were relatively high and long-lasting. The
muscle activity correlated with the catching/throwing
of the medicine ball and the pulling of the load in
the phase “Cable pullover”. The comparison of the
measured values of the Balance Pad and Board dur-
ing the third phase showed that the average values for
this subject were higher when using the Pad. How-
ever, the recorded impression of the subject showed
the highest effort during the second phase. The cate-
gorization based on figure 3 generated the following
ranking for the level of difficulty (lowest first): Bal-
ance Pad - Balance Board - Ortho Pad.
Figure 3: Comparison of the three exercisers.
The heart rate is used as an indicator for stress.
Figure 4 shows the average heart rate during the five
phases for each of the three groups. The test persons
of group one have the lowest heart rate. In contrast
to the other groups, the subjects of group one were
familiar with the measuring instruments. The higher
heart rate during the initial phase could have been
caused by the tension of the subjects. This assump-
tion is supported by the course of the heart rate of
group two which is lower during the last phase than
in the first one. The recorded course of the heart rate
for group one follows a similar pattern than the one
for group three, both increasing during the first four
phases. In contrast, the heart rate for group two al-
ready decreases after the second phase. This confirms
that training causes a learning effect and leads to the
subjects starting to feel comfortable after an initial
phase of adaptation.
The test persons’ subjective impressions when us-
ing the “Ortho Pad” are documented in figure 5.
When analyzing the heart rate, the lowest stress level
was recorded for group two while group 3 had the
highest one. Comparing the test persons’ subjective
impressions (figure 5) with the measured heart rate
InvestigationoftheSensorimotorTrainingusingWirelessSensorNetworks-AnalyzingThreeDifferentExerciser
415
Figure 4: Comparison of the heart rate at the Ortho Pad.
Figure 5: Subjective impressions after using the Ortho Pad.
(figure 4), only group three showed a direct link be-
tween the two.
4 DISCUSSION
The main finding of this study is that the correct appli-
cation of an exerciser depends on several factors. The
analysis of the EMG patterns from one subject using
the three exercisers shows that the patterns are dif-
ferent with respect to the produced voltage values as
well as the activation duration of each muscle. Hence,
the analysis of the individual exerciser could be a tool
for the optimal therapy planning. Despite identical
tasks, the muscle activity patterns of the individual
subjects differ strongly. This implies that the course
of the muscle activity is also depending on the sub-
jects’ skills. Therefore, it is recommended to analyze
the individual behavior of a patient at the various ex-
ercisers in order to plan his or her therapy effectively.
One possible application for the sensorimotor
training is the fall prevention, especially for older pa-
tients. Depending on the physical fitness of these pa-
tients, a verification of the general effort could be use-
ful.
The study further revealed a limited correlation
between the subjective impressions of the test per-
sons and the measured values. Therefore, the thera-
pist should critically review the patient’s comments.
5 CONCLUSIONS
The evaluation of the exercisers showed that differ-
ent subjects can have different courses of movement
when using the same equipment. Hence, mobile sen-
sors could greatly improve the study of the individ-
ual interaction between a subject and the equipment
in use. Furthermore, the sensors can assist with ther-
apy documentation in general as well as the progress
recording of a particular muscle.
A possible continuation of this study could be the
measurement and analysis of the gyroscope data. In
combination with the muscle activity pattern, the data
can offer additional information about the way a pa-
tient is reacting to movements of the exerciser. Within
this context, the calculation of the reaction time of the
muscle could also be considered. Hence, it is possi-
ble to characterize the individual muscles needed to
improve for example the instability of a joint.
REFERENCES
Dohm-Acker, M., Spitzenpfeil, P., and Hartmann, U.
(2008). Auswirkung propriozeptiver trainingsgerte
auf beteiligte muskulatur im einbeinstand. Sportverl
Sportschad, pages 52–57.
H
¨
afelinger, U. and Schuba, V. (2010). Koordinationsthera-
pie: Propriozeptives Training. Meyer & Meyer Ver-
lag.
Husar, P. (2010). Biosignalverarbeitung. Springer.
Lukas, C., Fr
¨
ohlich, V., Kapferer, C., and Zelder, C. (2011).
Sprunggelenksverletzungen im Basketball: Hinter-
grnde, Therapie und Prophylaxe. Books on Demand.
Merletti, R. and Parker, P. A. (2004). Electromyography.
John Wiley & Sons.
SENIAM project (2012). Sensor placement. Website.
Available online at http:// www.seniam.org; visited on
October 25th 2012.
Shimmer Research Support (2012a). ECG User Guide Rev
1.1. Shimmer Research.
Shimmer Research Support (2012b). EMG User Guide Rev
1.2. Shimmer Research.
Sport-Thieme (2012a). Airex
R
balance-pad. Web-
site. Available online at http://www.webcitation.org/
6BgijgPOx; visited on October 25th 2012.
Sport-Thieme (2012b). Ortho-pad. Website. Available on-
line at http://www.webcitation.org/6Bgiy6qH5; vis-
ited on October 25th 2012.
Sport-Thieme (2012c). Sport-thieme
R
sport- und ther-
apiekreisel. Website. Available online at http://
www.webcitation.org/6BgicOk7Y; visited on October
25th 2012.
HEALTHINF2013-InternationalConferenceonHealthInformatics
416