This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright
Author's personal copy Journal of Bodywork & Movement Therapies (2011) 15, 453e464
available at www.sciencedirect.com
journal homepage: www.elsevier.com/jbmt
ASSESSMENT TECHNOLOGY
Real-time stability measurement system for postural control Alpha Agape Gopalai, B. Eng (Mechatronics) Hons a, S.M.N. Arosha Senanayake, M. Eng, PhD a,*, Loo Chu Kiong, PhD, B.Eng (Mechanical) Hons b, Darwin Gouwanda, B. Eng (Mechatronics) Hons a a School of Engineering, Monash University Sunway campus, 2, Jalan Lagoon Selatan, 46150 Bandar Sunway, Selangor, Malaysia b Centre for Artificial Intelligence and Robotics, Multimedia University, Malacca, Malaysia
Received 25 June 2010; received in revised form 29 September 2010; accepted 19 October 2010
KEYWORDS Weighted center of pressure; Pressure concentrations; Instrumented platform; Postural response; Perturbed surface
Summary A method for assessing balance, which was sensitive to changes in the postural control system is presented. This paper describes the implementation of a force-sensing platform, with force sensing resistors as the sensing element. The platform is capable of measuring destabilized postural perturbations in dynamic and static postural conditions. Besides providing real-time qualitative assessment, the platform quantifies the postural control of the subjects. This is done by evaluating the weighted center of applied pressure distribution over time. The objective of this research was to establish the feasibility of using the forcesensing platform to test and gauge the postural control of individuals. Tests were conducted in Eye Open and Eye Close states on Flat Ground (static condition) and the balance trainer (dynamic condition). It was observed that the designed platform was able to gauge the sway experienced by the body when subject’s states and conditions changed. ª 2010 Elsevier Ltd. All rights reserved.
Introduction
* Corresponding author. Tel.: þ60 35514 6249; fax: þ60 35514 6001. E-mail addresses:
[email protected] (A.A. Gopalai),
[email protected] (S.M.N.A. Senanayake),
[email protected] (L.C. Kiong),
[email protected] (D. Gouwanda).
The essence of human motion is the maintaining of postural alignment while standing upright. Postural alignment affects postural control as it controls the amount of effort required by the body for support, against gravity (Kantor et al., 2001). The human body incorporates various mechanisms to achieve postural alignment, which is mainly influenced by the availability and validity of the visual and the somatosensory receptors. The visual and somatosensory mechanism
1360-8592/$ - see front matter ª 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbmt.2010.10.005
Author's personal copy 454 functions as inputs to the Central Nervous System (CNS) (Danis et al., 1998). The CNS provides referential context for postural adjustments, necessary to control the body balance (Raymakers et al., 2005). The maintenance of control in body balance together with postural alignment for stability and steadiness are crucial aspects to injury avoidance (Page, 2006). Over the years many tools and methodologies have been developed to study postural alignment and postural control of the human body in various conditions (Chaudhry et al., 2008). Stability and steadiness are often assessed as the amount of displacement made by the Center of Pressure (Danis et al., 1998; Nault et al., 2002; Allard et al., 2004; Dalleau et al., 2007). There are a number of equipments that clinicians use to gauge balance of individuals. The most common equipments were reviewed and discussed in Chaudhry et al., (2008). These included Force Plates, Balance Master and Equitest. Chaudhry et al., (2008) concluded the Balance Master and Equitest had the most relevance to daily life, due to its moveable support surface that emulates real life conditions. Other studies on balance had also reached a similar conclusion of moveable platforms (Allum et al., 2010; Carpenter et al., 1999; Bloem et al., 1998). The drawbacks of the equipments presented in Chaudhry et al., (2008) however, were the high cost involved in acquiring them and their large space requirements. Such equipment is not designed to be mobile, hence limiting its application to a clinical or laboratory setting. This work presents an alternative method to assess stability and steadiness. The alternative method involves the design of an instrumented force-sensing platform that measures the displacement in weighted center of the applied pressure between the human feet and the platform’s surface. These points are tracked over time to facilitate understanding of migration of the weighted center of applied pressure. The surface of the platform was instrumented with arrays of Force Sensing Resistors (FSR). Common parameters used for assessing postural control are the sway velocity, sway area and sway path. The authors in Siqueira et al., (2009) and Mann et al., (2010) measured sway velocities in order to assess postural control of subjects. A different approach was taken in Owen et al., (1998) which used the mean sway path to determine the deterioration of postural control in the presence of fatigue. The parameter measured in this work to assess postural control was the sway area, as was previously reported in Nardone et al., (1997) and ReedJones et al., (2008). Equipment that is commonly used to measure dynamic condition balance, such as the Balance Master and Equitest utilize dual force plates to individually measure the required parameters of each foot. Unlike the force sensing platform that uses FSR as its sensing element, the force plate utilizes load cell(s) to measure the vertical component in the geometric center of the platform. Load cells measure the average exerted force in a specific area covered. Due to its construction, single force plates are not able to provide parallel feet assessment, and would require an additional force plate for such measurement (one force plate per feet), as was reported in Jancova (2008). This work proposes a single force-sensing platform for parallel
A.A. Gopalai et al. foot assessment in determining the weighted center of applied pressure migration. The aim of this investigation was to study the feasibility of using FSRs as a basic sensing element to provide observers an alternative method, for qualitative measure of postural control and proprioceptive strength in real-time. The qualitative measure of proprioceptive strength was then used to describe the postural control (in static and dynamic states) for Eye Open (EO) and Eye Closed (EC) conditions (Siqueira et al., 2009). The platform presented in this work, addressed the cost constraint that was associated with current technologies (Chaudhry et al., 2008), while being portable in nature, allowing for on-site tests to be conducted.
Materials and methods The designed system, possesses two functionalities depending on its utilization, real-time or post-acquisition. Figure 1 briefly summarizes these two functionalities and the processes involved to obtain the final results. The processes are ordered according to the chronology of processes in the work.
Subjects The subjects for this study consisted of 18 healthy subjects (9 Males and 9 Females), volunteers from the community, aged between 20 and 30 years. All participants were healthy and had no known neurological or muscular disorder. The subject group had the following average readings, age Z 23.69 years (S.D. Z 2.39), mass Z 62.49 kg (S.D. Z 10.67), height Z 167.25 cm (S.D. Z 7.59) and Body Mass Index (BMI) Z 22.17 kg/m2 (S.D. Z 2.14). Before any readings were taken, the researcher went through in detail the entire purpose and procedure of the experiment. Subjects were informed on the potential hazards that may be involved during data collection. All subjects gave their informed consent to participate in this study. Safety hand rails were introduced to assist subjects mount the balance trainer, and served as support in event subjects lose balance during data acquisition, Figure 2. The tests were conducted in the laboratory for BioInspired Robotic Devices (BIRDs) of the School of Engineering, Monash University, Sunway campus with the ethic clearance from the Monash University Human Research Ethics Committee (MUHREC) (CF10/0762e2010000339).
Instrumented platform The instrumented platform is a force-sensing platform. The constructed platform was based on earlier concepts of a force-sensing device presented in Senanayake et al., (2007); Khoo et al., (2007), and Gouwanda and Senanayake, (2008). These studies utilized FSRs which are a polymer thick film device that exhibits a decrease in resistance with an increase in the force applied to the active surface, Figure 3(a). The change in resistance by these sensing elements were converted to voltage linearly. A platform of this nature must have adequate sampling rate, to ensure a complete profile capture during data acquisition.
Author's personal copy Real-time postural stability measurement system
455 sampled by the host computer. Figure 3(b) illustrates the architecture of circuits implemented with the force sensing platform and how they relate to each other.
BOSU balance training platform The BOSU Balance Trainer is a unique balance, core stability and proprioception training device that is widely used by professionals and national athletes to help achieve body balance rapidly and safely. The platform provides an unstable balance surface when subjects stand on it. The force sensing platform was attached to the surface of the BOSU Balance Trainer, Figure 4. The force sensing platform had negligible movement over the BOSU surface because the platform was hinged to supports on the surface. The attachment of the force sensing platform allowed for the observation and evaluation of human postural control in dynamic conditions. The BOSU surface measured 635 mm in diameter (Figure 4(a)) and had a variable height, x mm (Figure 4(c)), depending on the amount of air introduced into its inflatable chamber. The perturbation experienced by the subjects once on the platform was not generated by electrical motors, controlled by computers or a program. The direction and angles of perturbation were solely dependent on the subjects’ body sway (i.e. perturbations were self inflicted). The balance trainer tilted in the direction of the net forces acting on the surface of the balance trainer. Figure 5 illustrates measurements taken by a gyroscope placed on board the balance trainer. The graphs depict the maximum and minimum amount of orientation that the subject may be subjected to whilst on the platform (z40 in both planes, Anterior-Posterior and Medial-Lateral). The introduction of the BOSU balance trainer allowed for investigation of immediate defensive postural reactions and the adaptation of postural control mechanisms when presented with an oscillatory surface. Figure 1 General overview of the process flow in the work presented.
Experimental procedures
The minimum sampling frequency for a platform of this nature is 50 Hz (Gouwanda and Senanayake, 2008), the designed platform has sampling rates of up to 200 Hz (all experiments were conducted at a sampling rate of 200 Hz). The designed force-sensing platform in this study measured 635 mm in diameter. This diameter was selected to ensure the largest possible instrumentation area allowing for natural foot positioning of subjects, ensuring readings acquired reflected naturally pathological conditions (Reed-Jones et al., 2008; Amiridis et al., 2003; Prieto et al., 1996). The platform was fitted with 122 units of FSR, arranged in a matrix form. Each FSR has a sensing area of 126.68 mm2 and was placed 40 mm apart (from the center of each sensing element). Raw signals from the sensors were then passed through a signal conditioning circuit in real-time before any further analysis. The FSRs were connected to non-inverting operational amplifier, for amplification of signal strength. Outputs from the operational amplifiers were in turn connected to a switching circuit before being
There have been many studies conducted in order to measure and to quantify human body sway under various conditions and environments (Blaszczyk et al., 2009; Pascolo et al., 2009; Lamoth et al., 2009). The common method of investigation was, the ability of the subjects to maintain body balance and posture in EO and EC states. The experimental procedures discussed in this paper adapts similar procedures established in Blaszczyk et al. (2009), Pascolo et al. (2009) and Lamoth et al. (2009). All acquisitions were captured in barefoot conditions, to ensure that the force profile readings acquired were not averaged due to the soles of the subjects’ footwear (Urry, 1999) and to provide sensorimotor maximum amount of appropriate afferent information (Page, 2006). Visual feedback was provided to end-users for observation of force concentration in real-time. The subject’s vertical projection of force (while on the force sensing platform) was recorded for both states (EO and EC) in both conditions (static and dynamic), and was stored on board a computer. The host computer sampled the instrumented platform via a Data Acquisition (DAQ) card (National Instrument’s PCIMIO-16E-1).
Author's personal copy 456
A.A. Gopalai et al.
Figure 4 (a)e(d) Technical drawing of the balance trainer (e) the force sensing platform mounted on the balance trainer for acquisition of readings to simulate perturbed conditions.
was called static condition because the base of support on which the subjects stood, was firm.
Figure 2 Subject mounted on the balance trainer fitted with the force-sensing platform and hand rails for additional safety support in dynamic condition.
Flat groundestatic condition Subjects were required to maintain body balance for 30 s in EO while standing on the force sensing platform placed on flat ground followed by a second set of readings for 30 s, in EC. This arrangement was alternated between acquisitions, for 3 pairs of EO and EC readings. This served as the basal data for subjects. Subjects were allowed to take breaks in between each pair of acquisition, if required. The method
Perturbed surfaceedynamic condition Subjects were required to mount the balance trainer fitted with the force-sensing platform (on the balance trainer’s surface) for 30 s in EO, followed by 30 s, in EC. The base of the balance trainer is convex, designed to challenge postural control of subjects, Figure 2. Breaks were introduced between each pair of readings (EO and EC) before resuming with the next pair of readings, in order to eliminate the effect of fatigue on postural sway (Nardone et al., 1997, p309). Data was logged for 3 pairs of readings. This section of the experiment was termed dynamic due to the progression of instability of the base of support. The base of support was dependent on the subjects’ body sway (perturbations are self inflicted). The balance trainer tilts in the direction of the net force exerted on its surface.
Data analysis In static and dynamic conditions, subjects’ force profile was acquired from the FSRs. Signals from FSRs were sampled via
Figure 3 (a) Force Sensing Resistors (FSR), exhibits a decreases in resistance with an increase in the force applied to its sensing area/surface. (b) Architecture of circuits for force sensing platform, beginning at the sensing element (FSR) to the sampling process by the host computer.
Author's personal copy Real-time postural stability measurement system
Figure 5
457
Range of perturbation induced by the balance trainer (a) in the M/L plane and (b) in the A/P plane.
the DAQ Card and passed through a moving averaging filter. The moving average filter technique was used to smooth out high frequency fluctuations that may be present in the sampled data due to electrical noise. This filtering technique was specifically chosen for its simplicity which reduces execution time, which is a crucial feature in a realtime feedback application. The acquired force profiles were displayed visually in real-time for qualitative assessments. The readings were also stored on board the host computer for post-acquisition analysis. Stored data sets can be retrieved for further analysis to quantify postural control. The proprioceptive strength of an individual is demonstrated by the ability to maintain the body posture in its ‘neutral’ position. An individual with good proprioceptive strength demonstrates a well controlled displacement of the COP for both EO and EC trials in static and dynamic conditions (Jancova, 2008). Visual representation of data The real-time output of the force-sensing platform was formatted according to the rainbow color scale. A rainbow color scale represents low force intensities with colors closer to black (cold colors), while high force intensities with colors closer to white (hot colors). This form of real-time feedback to end-users eases the identification of regions with high force concentrations by the foot, Figure 6(a). For data representation in post-acquisition, the data set was first checked for redundancy. Redundancy in the data sets referred to regions of the platform that did not come in contact with the subjects foot. A simple thresholding algorithm was applied to the acquired data set to isolate the areas of the redundant data points using equations (1)e(4). Since the force-sensing platform consists of FSRs arranged in a matrix form, each sensing unit can be individually identified as elements of a matrix. Let m be the total number of rows and n be the total number of columns, while the sensor matrix is represented in the data set as FSR and T a pre-determined threshold value. Figure 6(b) depicts the final result of applying equations (1)e(4), the boundary of the foot can now be plotted for ease of visualization.
m X In Z FSRði; : Þ
ð1Þ
iZ1 n X Im Z FSRð :; iÞ
ð2Þ
iZ1
fIn : Im TgðRegion is in contactÞ
ð3Þ
fIn : Im < TgðRegion is not contactÞ
ð4Þ
Identifying pressure concentration regions During the real-time feedback monitoring process (to the end-user), pressure concentrations were easily detected as depicted in Figure 6(a) (Qualitative method). These pressure concentration regions do not provide substantial quantitative relationship between balance control and foot pressure details. Hence, post-processing of the acquired data in its time series is required, to provide end users with a quantitative measure of the subject’s balance and postural control. Regions with high concentration of pressure collected over time are of interest in this study
Figure 6 Foot boundaries (a) Real-time feedback snapshot, depicting regions of high force concentrations (b) Reproduced foot boundary from reduced data set indicating regions of weighted center of the applied pressure over time.
Author's personal copy 458
A.A. Gopalai et al. Figure 7, depicts a portion of the force sensing platform (consisting of four FSRs arranged in a two-by-two matrix). The FSRs are distributed symmetrically in a similar manner across the force sensing platform. Equations (5) and (6) were applied across the force sensing platform recursively. At the end of the calculations, the acquired force profile was represented by FY , which is the total force acting perpendicularly on the foot at coordinates (X,Z ) per data sweep. A data sweep refers to the sampling of all 122 FSRs in a single clock cycle.
Figure 7 Magnitude and location of the resultant force FY calculated from the combined signals of the FSRs.
because it shows exactly which part of the foot the subject was using to maintain the ‘neutral’ position while standing, Figure 6(b). In this work, the pressure concentrations were calculated as the weighted center of applied pressure, which is referred to in this work as weighted center of pressure (COP). The COP of each foot was calculated using (5) and (6) to obtain the COP along the X-axis (Medial-Lateral, ML) and the Z-axis (Anterior-Posterior, AP). The equations in (5) and (6) results in the physical location (1 unit corresponds to 40 mm on the platform) of the weighted center of applied pressure (X,Z ), per data sweep. To obtain the time series of the migration of the COP this calculation was performed on the entire data sweep.
x ðFSRXO þ FSRXZ Þ ðFSROO þ FSROZ Þ XZ 1 þ 2 FY
ð5Þ
z ðFSROZ þ FSRXZ Þ ðFSROO þ FSRXO Þ ZZ 1 þ 2 FY
ð6Þ
FY ZFSROO þ FSRXO þ FSROZ þ FSRXZ
ð7Þ
The measure of postural control The mean value of the subject’s COP was calculated for each plane, to determine the range of postural sway experienced by the subject. Readings acquired from the subjects were averaged across the three collected sets, the standard deviations were also obtained and tabulated. The sample set was accepted as a valid COP reading if and only if the calculated value was within 95% interval (1.97 standard
Figure 8 Screen shots of of real-time readings from the platform at certain point in time (a) EO state in static condition (b) EC state in static condition (c) EO state in dynamic condition and (d) EC state in dynamic condition.
Author's personal copy Real-time postural stability measurement system
Figure 9
459
Shift in COP on flat ground (a) Left Foot in EO (b) Right Foot in EO (c) Left Foot in EC (d) Right Foot in EC.
deviations - normal distribution assumed). The maximum and minimum values along the X and Z axes for each foot were identified. The area which contained 95% of the COP distribution was obtained. The area of this distribution was a key indicator of the subject’s postural control and proprioceptive strength. Subjects who demonstrated a large distribution area, tend to have a poorer postural control, distribution area and postural control are inversely proportional to each other (Lamoth et al., 2009; Jancova, 2008; Dalleau et al., 2007).
Results and analysis Real-time data representation The rainbow color scale visualizes the real-time output of the force-sensing platform. This method of representation provided real-time visual qualitative assessment of subjects in static and dynamic conditions. Figure 8 shows an example of the real-time feedback observed by users. End-
users can easily identify pressure points on the subject’s feet (white regions). This was due to the proprioceptors at the foot keeping the body in its ‘neutral’ position (Winter et al., 2003).
Analysis of postural Response in static conditions Figure 9(a) and (b) illustrates the shift in the calculated COP in EO on flat ground. Figure 9(c) and (d) illustrates the shift in the calculated COP in EC on flat ground. Results of the tests for static condition in EO were tabulated in Table 1 while the results for EC were tabulated in Table 2. The AP and ML measures in the tables are in terms of units of the force-sensing platform. The area of distribution was calculated based on the range of AP and ML of the left and right foot.
Analysis of postural Response in dynamic conditions This section of the results looks at the balancing capabilities of the subjects in dynamic conditions. Figure 10(a) and (b) illustrates the shift in the calculated COP in EO on the
Author's personal copy 460
A.A. Gopalai et al.
Table 1 Range of calculated COP per foot for each subject and the average area of distribution, in static condition for EO conditions. Sub
Left EO ML
F1 F2 F3 F4 F5 F6 F7 F8 F9 M1 M2 M3 M4 M5 M6 M7 M8 M9
2.49 3.78 4.24 4.67 3.93 4.79 3.36 4.84 5.27 3.71 2.66 4.14 4.27 3.74 3.35 3.32 4.50 4.99
AP
0.005 0.013 0.008 0.022 0.016 0.042 0.040 0.009 0.013 0.005 0.008 0.021 0.041 0.042 0.029 0.065 0.048 0.099
Area(mm2)
Right EO
7.13 6.00 7.62 8.03 7.27 6.74 5.95 7.93 9.10 7.93 5.88 6.83 7.51 5.75 6.00 5.99 6.53 6.31
ML
0.016 0.030 0.018 0.030 0.052 0.067 0.086 0.148 0.053 0.014 0.025 0.039 0.042 0.044 0.051 0.029 0.081 0.111
balance trainer that were introduced to emulate dynamic conditions. Figure 10(c) and (d) illustrates the shift in the calculated COP in EC on a movable platform. Results of the tests for dynamic conditions in EO was tabulated in Table 3 while the results for EC was tabulated in Table 4. The AP and ML measures in the tables are in units of the force sensing platform.
8.38 9.46 9.30 9.78 9.30 9.88 9.28 9.76 8.63 9.40 8.75 9.64 9.66 9.30 9.67 8.84 9.48 9.40
AP
0.012 0.007 0.405 0.019 0.041 0.020 0.033 0.079 0.084 0.006 0.013 0.014 0.018 0.007 0.025 0.015 0.008 0.018
6.29 5.50 7.99 6.25 7.99 5.00 5.40 6.01 6.07 6.58 5.41 5.77 6.41 5.01 6.54 4.53 5.90 5.04
0.026 0.053 0.074 0.152 0.074 0.108 0.118 0.134 0.239 0.036 0.032 0.066 0.015 0.027 0.028 0.025 0.067 0.017
9.99 19.94 80.66 90.93 98.28 127.18 185.89 303.80 529.05 6.92 15.17 44.62 51.02 52.01 55.31 58.81 112.97 290.83
Discussion The shifts observed in pressure distribution from the graph for standing in static conditions (EO and EC) oscillates within a small concentrated region Figure 9. The EO readings served as basal test conditions for comparison against readings in dynamic conditions. The average area of sway
Table 2 Range of calculated COP per foot for each subject and the average area of distribution, in static condition for EC conditions. Sub
Left EO ML
F1 F2 F7 F6 F5 F3 F4 F8 F9 M2 M1 M5 M6 M7 M3 M8 M4 M9
2.50 3.63 3.39 4.59 4.04 4.20 4.71 5.03 5.30 3.71 2.66 4.27 3.74 3.35 3.80 3.32 4.50 6.56
AP
0.008 0.066 0.054 0.018 0.073 0.012 0.066 0.018 0.020 0.027 0.006 0.057 0.037 0.087 0.059 0.116 0.445 0.386
Area(mm2)
Right EO
7.16 5.98 5.98 6.43 6.98 7.56 8.07 7.61 9.20 5.88 7.63 5.82 6.03 5.94 6.58 6.55 7.53 6.09
ML
0.019 0.042 0.098 0.040 0.244 0.025 0.091 0.148 0.067 0.021 0.011 0.050 0.027 0.087 0.058 0.184 0.045 0.074
8.87 9.50 9.28 10.00 9.25 8.76 9.74 9.80 9.18 8.76 9.53 9.32 9.68 9.06 9.49 9.48 9.63 9.34
AP
0.018 0.037 0.049 0.020 0.034 0.010 0.080 0.190 0.116 0.010 0.032 0.011 0.017 0.017 0.038 0.015 0.071 0.017
6.22 5.51 5.37 5.59 8.00 6.36 6.32 6.23 5.94 5.43 6.31 4.95 6.74 4.46 5.70 5.92 6.41 5.03
0.026 0.136 0.124 0.606 0.163 0.270 0.335 0.163 0.283 0.012 0.060 0.041 0.241 0.093 0.212 0.102 0.050 0.068
15.57 198.87 290.52 326.84 599.53 705.56 839.10 858.07 871.83 17.76 50.87 83.35 129.76 234.48 293.83 586.44 600.82 764.78
Author's personal copy Real-time postural stability measurement system
461
Figure 10 Shift in COP on balance trainer (dynamic condition) for the entire data acquisition (a) Left Foot in EO (b) Right Foot in EO (c) Left Foot in EC (d) Right Foot in EC.
calculated for the subject group in static EO conditions, matches previously published data (Raymakers et al., 2005) for subjects of the same age group (young and healthy). It has been established that humans experience a poorer postural control in absence of a valid visual cue (Kaesler et al., 2007; Siqueira et al., 2009). The platform was able to detect the change in postural sway experienced by the body, in absence of a valid visual cue in static conditions. This was observed in the static condition experiments when subjects recorded a larger distribution area in EC, Table 2, as compared to the EO, Table 1. Analysis on the data from the force sensing platform in static condition was also able to localize the regions of high concentration of the COP. In the EO experiment the distribution sites were observed to be within a single region, Figure 9(a) and (b). However, in the EC experiment there were two major concentration sites in which the calculated COP oscillates within Figure 9(c) and (d). This observation was in parallel with the observation that in absence of a visual cue, subjects become solely reliant on the
proprioceptors to inform the CNS on the body’s position with respect to its surrounding. The proprioceptors now become responsible for informing the CNS when the body is swaying excessively in a certain direction, so that compensatory measures can be made to counter the sway experienced. In the experiments for dynamic condition, subjects recorded a much higher oscillation and larger distribution of COP, for both legs, Figure 10. This was due to the reason that the moveable platform used to emulate dynamic condition tilts in the direction of the nett force exerted on its surface. This tilting to one direction required compensatory measures, which was exerted by the opposite portion of the foot, to maintain ‘neutral’ position (controlled by the CNS). This sway of forward and backward or left to right motion, is the reason for the overall increase in oscillation and distribution area of the COP in dynamic conditions. Despite being distributed over a larger area, the oscillation observed for the EO is centered within a single area Figure 10(a) and (b). The data in Table 3 also identifies that subjects tend to depend on one foot. This
Author's personal copy 462
A.A. Gopalai et al.
Table 3 Range of calculated COP per foot for each subject and the average area of distribution, in dynamic condition for EO conditions. Sub
Left EO ML
F1 F2 F5 F3 F4 F7 F6 F8 F9 M2 M3 M1 M5 M4 M9 M6 M8 M7
3.69 3.15 3.02 3.46 3.82 3.66 2.87 3.38 3.35 3.45 3.50 4.49 3.76 2.67 3.79 3.35 3.32 3.82
AP
0.033 0.035 0.106 0.110 0.234 0.388 0.172 0.076 0.044 0.129 0.080 0.128 0.035 0.130 0.040 0.152 0.198 0.230
Area(mm2)
Right EO
6.40 6.43 7.46 6.54 7.52 7.28 6.97 6.44 7.06 6.76 6.49 7.77 6.24 6.79 6.12 6.65 7.25 7.28
ML
0.090 0.253 0.055 0.082 0.129 0.061 0.108 0.074 0.031 0.018 0.193 0.058 0.253 0.071 0.269 0.154 0.175 0.242
was observed by a significantly smaller overall distribution of the individual foot area of distribution, which indicates subjects tend to rest their body weight on one foot while trying to compensate for the sway experienced with the opposite foot. In EC for dynamic condition, subjects become heavily reliant on the proprioceptors at the feet, and subjected to significantly higher body sways, Table 4. Unlike the previous observations, there are no clear observable distribution sites in EC for dynamic conditions. Subjects were observed to rest their body weight on a particular foot, as was in EO
9.43 9.75 9.36 10.48 8.95 10.53 9.78 9.62 9.68 9.46 8.85 9.89 9.33 9.02 9.33 9.53 9.53 9.45
AP
0.038 0.023 0.176 0.164 0.089 0.124 0.234 0.566 0.289 0.095 0.030 0.067 0.020 0.036 0.021 0.099 0.084 0.122
6.72 6.60 6.55 6.98 6.78 6.75 6.37 6.67 5.83 6.53 5.97 6.56 7.21 5.88 7.07 6.11 5.57 5.64
0.321 0.390 0.193 0.210 0.177 0.265 0.171 0.173 0.365 0.098 0.056 0.159 0.493 0.351 0.623 0.099 0.068 0.335
386.23 453.25 1017.98 1113.06 1173.04 1450.77 1514.68 2648.73 2731.17 296.14 439.04 460.22 486.31 562.07 610.66 849.74 1028.85 2464.90
for dynamic conditions. This was observed graphically in Figure 10(d) when a particular foot graph of the subject has a smaller concentrated distribution as opposed to the opposing foot in Figure 10(c). Subjects who could not maintain balance and were at the verge of falling due to the lost of balance, came in contact with the safety hand rail. In such cases, the acquisition period was cut short. Portions of the data set after contact with the safety hand rail could no longer be considered valid because it has been corrupted by additional sensory information that was unintended during the data acquisition. Out of the 18
Table 4 Range of calculated COP per foot for each subject and the average area of distribution, in dynamic condition for EC conditions. Sub
Left EO ML
F1 F2 F7 F5 F4 F6 F9 F8 F3 M2 M3 M4 M5 M1 M7 M8 M6 M9
3.69 3.15 3.66 3.02 3.82 2.87 3.35 3.38 3.46 3.45 3.46 2.66 3.76 4.49 3.82 3.32 3.35 2.64
AP
0.315 0.329 0.230 0.237 0.249 0.375 0.232 0.411 0.566 0.122 0.168 0.257 0.089 0.357 0.610 0.386 0.338 0.722
Area(mm2)
Right EO
6.40 6.43 7.27 7.46 7.52 6.97 7.06 6.44 6.54 6.76 6.65 6.79 6.24 7.77 7.28 7.25 6.65 5.72
ML
0.182 0.199 0.277 0.178 0.215 0.267 0.280 0.294 0.245 0.055 0.207 0.085 0.220 0.297 0.212 0.156 0.347 0.181
9.43 9.75 10.53 9.36 8.94 9.78 9.68 9.62 10.05 9.46 8.83 9.02 9.33 9.89 9.45 9.53 9.53 6.75
AP
0.063 0.121 0.142 0.336 0.339 0.260 0.374 0.241 0.573 0.148 0.037 0.154 0.086 0.141 0.120 0.255 0.133 1.937
6.72 6.60 6.75 6.55 6.77 6.37 5.83 6.67 6.98 6.53 5.97 5.88 7.21 6.56 5.64 5.57 6.11 8.13
0.331 0.346 0.399 0.368 0.348 0.314 0.364 0.344 0.369 0.115 0.041 0.126 0.527 0.219 0.206 0.372 0.367 1.517
2000.60 2739.75 3083.48 4249.04 4379.42 4650.31 5144.93 5218.94 8967.69 606.33 926.04 1057.55 1660.04 3506.28 3942.71 3970.52 4237.87 78,580.89
Author's personal copy Real-time postural stability measurement system subjects that performed the trial, only 2 subjects found it impossible to maintain balance on the balance trainer in EC.
Conclusion The results obtained demonstrate the force-sensing platform’s ability to test and gauge postural control by means of proprioceptive strength. The research concentrated on the means of monitoring the proprioceptive strength by utilizing the FSR to monitor high pressure concentration sites. The output of the system enables end users to easily identify areas of the subjects’ foot that experiences a high level of forces, in real-time (static and dynamic conditions). The high level of forces within a certain area, indicates the region of the foot that subjects tend to use to balance themselves. Training and feedback helps address and target the identified area in which force concentration takes place during balancing and helps reduce pressure concentration. The outputs from the post-processing procedures provides end-users with a quantitative assessment of the postural control by means of proprioceptive control. Good proprioception is seen as the ability to maintain balance and a relatively constant pressure throughout the surface of both feet. Such features of pressure distribution at the feet is desirable, for a lower risk of foot related injuries. The designed system helps monitor the proprioception at the feet and can be used to identify individuals that begin to demonstrate early signs of balance deterioration. In a clinical setting, balance deterioration would have a different impact for athletes and non-athletes. A balance deterioration to an athlete would mean a drop in performance or an increased probability for sprains and foot related injuries. Deterioration to a non-athlete would indicate an increased risk of falls in daily activity, which could be fatal depending on the age group of the individual. Based on the results obtained, which were in agreement with other literature, the use of the FSR as a basic sensing element to provide quantitative and qualitative measurements of postural control was validated. The method utilized a force-sensing platform to detect the location of the weighted center of applied pressure per data sweep and analyzed the readings acquired with respect to time. The calculated coordinates were used to describe the migration of the weighted centers over time. The presented method to assess balance, was shown to be sensitive to changes in the postural control system and demonstrated itself to be a reliable device to measure proprioceptive strength of individuals. This work can be further expanded to test the effects of individual’s lifestyle or fitness level on postural control, within a similar age group. Further improvements can be introduced to the signal conditioning and switching circuit. The implementation of a micro-controller to replace these circuits will improve the performance of the platform and allow for higher sampling rates to be achieved.
Conflict of interest None.
463
Acknowledgement This work was supported by Monash University, Sunway Campus, The Ministry of Science, Technology and Innovation (MOSTI), Malaysia, and Moves International Fitness.
References Allard, P., Chavet, P., Barbier, F., Gatto, L., Labelle, H., Sadeghi, H., 2004. Effect of body morphology on standing balance in adolescent idiopathic scoliosis. Journal of Physical Medicine 83 (9), 689e697. Allum, J., Carpenter, M., Honegger, F., 2010. Directional aspects of balance corrections in man. IEEE Engineering in Medicine and Biology Magazine 22 (2), 37e47. Amiridis, G., Hatzitaki, V., Arabatzi, F., 2003. Age-induced modifications of static posture control in humans. Neuroscience Letters 350, 137e140. Blaszczyk, J.W., Cieslinska-Swider, J., Plewa, M., ZahorskaMarkiewicz, B., Markiewicz, A., 2009. Effects of excessive body weight on postural control. Journal of Biomechanics 42 (9), 1295e1300. Bloem, B.R., van Vugt, J.P.P., Beckley, D.J., Remler, M.P., Roos, R.A.C., Feb. 1998. Habituation of lower leg stretch responses in parkinson’s disease. Electroencephalography and Clinical Neurophysiology/Electromyography and Motor Control 109 (1), 73e77. Carpenter, M., Allum, J., Honegger, F., 1999. Directional sensitivity of stretch reflexes and balance corrections for normal subjects in the roll and pitch planes. Experimental Brain Research 129 (1), 93e113. Chaudhry, H., Bukiet, B., Ji, Z., Findley, T., 2008. Measurement of balance in computer posturography: Comparison of methodsea brief review. Journal of Bodywork and Movement Therapies in Press, Corrected Proof, e. Dalleau, G., Allard, M.S., Beaulieu, M., Rivard, C., Allard, P., 2007. Free moment contribution to quiet standing in able-bodied and scoliotic girls. European Spine Journal 16 (10), 1593e1599. Danis, C.G., Krebs, D.E., Gill-Body, K.M., Sahrmann, S., May 1998. Relationship between standing posture and stability. Physical Therapy 78 (5), 502. Gouwanda, D., Senanayake, S., 2008. Real time multi-sensory force sensing mat for sports biomechanics and human gait analysis. International Journal of Intelligent Systems and Technologies 3, 149e154. Jancova, J., 2008. Measuring the balance control systemereview. ACTA MEDICA 51 (3) 197e137. Kaesler, D., Mellifont, R., Kelly, P.S., Taaffe, D., Jan. 2007. A novel balance exercise program for postural stability in older adults: a pilot study. Journal of Bodywork and Movement Therapies 11 (1), 37e43. Kantor, E., Poupard, L., Bozec, S.L., Bouisset, S., Aug. 2001. Does body stability depend on postural chain mobility or stability area? Neuroscience Letters 308 (2), 128e132. Khoo, B., Senanayake, S., Gouwanda, D., Oct. 2007. A Portable Vertical Jump Analysis System. In: Fuss, F.K., Subic, A., Ujihashi, S. (Eds.), The Impact of Technology on Sport II. CRC Press, pp. 643e648. Lamoth, C.J., van Lummel, R.C., Beek, P.J., 2009. Athletic skill level is reflected in body sway: a test case for accelometry in combination with stochastic dynamics. Gait & Posture 29 (4), 546e551. Mann, L., Kleinpaul, J., Moro, A., Mota, C., Carpes, F., 2010. Effect of low back pain on postural stability in younger women: Influence of visual deprivation. Journal of Bodywork and Movement Therapies 14 (4), 361e366. Nardone, A., Tarantola, J., Giordano, A., Schieppati, M., 1997. Fatigue effects on body balance. Electroencephalography and
Author's personal copy 464 Clinical Neurophysiology/Electromyography and Motor Control 105 (4), 309e320. Nault, M., Allard, P., Hinse, S., Blanc, R.L., Caron, O., Labelle, H., Sadeghi, H., September 1, 2002. 2002. Relations between standing stability and body posture parameters in adolescent idiopathic scoliosis. Spine 27 (17), 1911e1917 [Miscellaneous article]. Owen, N., Leadbetter, A., Yardley, L., 1998. Relationship between postural control and motion sickness in healthy subjects. Brain Research Bulletin 47 (5), 471e474. Page, P., Jan. 2006. Sensorimotor training: a "global" approach for balance training. Journal of Bodywork and Movement Therapies 10 (1), 77e84. Pascolo, P.B., Carniel, R., Pinese, B., 2009. Human stability in the erect stance: alcohol effects and audio-visual perturbations. Journal of Biomechanics 42 (4), 504e509. Prieto, T., Myklebust, J., Hoffman, R., Lovett, E., Myklebust, B., 1996. Measures of postural steadiness: differences between healthy young and elderly adults. IEEE Transactions on Biomedical Engineering 43 (9), 956e966.
A.A. Gopalai et al. Raymakers, J.A., Samson, M.M., Verhaar, H.J.J., Jan. 2005. The assessment of body sway and the choice of the stability parameter(s). Gait & Posture 21 (1), 48e58. Reed-Jones, R., Vallis, L., Reed-Jones, J., Trick, L., 2008. The relationship between postural stability and virtual environment adaptation. Neuroscience Letters 435, 204e209. Senanayake, S., Looi, D., Liew, M., Wong, K., Hee, J., Oct. 2007. Smart Floor Design for Human Gait Analysis. In: Fuss, F.K., Subic, A., Ujihashi, S. (Eds.), The Impact of Technology on Sport II. CRC Press, pp. 189e192. Siqueira, C. M., Moya, G. B. L., Caffaro, R. R., Fu, C., Kohn, A. F., Amorim, C. F., Tanaka, C., 2009. Misalignment of the knees: Does it affect human stance stability. Journal of Bodywork and Movement Therapies in Press, Corrected Proof, e. Urry, S., 1999. Plantar pressure-measurement sensors. Measurement Science and Technology 10 (1), 16e32. Winter, D.A., Patla, A.E., Ishac, M., Gage, W.H., Feb. 2003. Motor mechanisms of balance during quiet standing. Journal of Electromyography and Kinesiology 13 (1), 49e56.
