Advances in rehabilitation technology using virtual reality and perturbations to assess and train gait and balance

Introduction

  The information about the effects of unstable footwear on human health has been increased significant in recent years. Round bottom feature of unstable structure footwear especially unstable in medial-Lateral direction can simulated the dynamic balance test conditions and providing adequate sensory motor stimulation to improved postural control[1]. Oxford Foot Model was used in this study, which will make a comparative experiment of foot kinematic during walking[2]. And try to find the difference between unstable structural soles and ordinary shoes.

Methods

  Twelve healthy female subjects walked through a 10-meter walk way wearing experimental shoes and control shoes respectively. Differences between the groups in triplanar motion of the forefoot, rearfoot and hallux during walking were evaluated using three-dimensional kinematic analysis adopted Vicon motion analysis system of 8-camera (Oxford Metrics Ltd., Oxford, UK) incorporating with Oxford Foot Model (OFM). Simple shoes are used as control shoes and the testing shoes are stuck rubber hemisphere to the outsole of simple shoes in the heel and forefoot (Fig1). Paired-samples T test was taken for comparison of the kinematics data, including the peak angle of joint (forefoot, rearfoot, hallux), standard deviation, the range of motion and significance set as p<0.05.

Fig.1 Lateral view of OFM maker placement and unstable elements

Results

Compare to the control group, subjects wearing unstable shoes demonstrated greater peak forefoot dorsiflexion, forefoot supination and longer time of hallux plantar-flexion in support phase. Additionally, subjects with unstable sole structure also demonstrated smaller peak forefoot plantar flexion, rearfoot dorsiflexion and range of rearfoot motion in sagittal and frontal plane. There is no significant difference between the groups in transverse plane. The unstable shoes with unstable structural elements leading to instability on foot motions. The difference mainly appeared in sagittal

and frontal plan (Fig2).

Fig.2 Average foot kinematics (°) during gait cycle

 

Discussion

A stimulation of unstable lead to the reinforcement of different flexion between middle and two ends of the foot model. The greater forefoot supination is inferred that unstable structure elements may has a control effect to strephexopodia people. The stimulation also reduce the range of rearfoot motion in sagittal and frontal planes to control the gravity center of the body and keep a steady state during walking.

 

Acknowledgement

The study was sponsored by ANTA Sports Products Co., Ltd.

 

References

1            S. Turbanski, H. Lohrer, T. Nauck, D. Schmidtbleicher, Training effects of two different unstable shoe constructions on postural control in static and dynamic testing situations, Phys. Ther. Sport. 12 (2011) 80–86. doi:10.1016/j.ptsp.2011.01.001.

2            H. Lohrer, S. Turbanski, T. Nauck, D. Schmidtbleicher, [Balance therapy shoes - a comparative analysis with respect to immediate training effects], Sport. Sport. 22 (2008) 191–195.

Introduction: Neurological damage due to a stroke (ST) or multiple sclerosis (MS) has a profound impact on patient’s balance. Although several tests are used to quantify balance impairment in neurological populations, few of them focus on weight-shifting abilities. This is of high relevance since most ST and MS patients have higher risk of falling than the normal ageing population, which is associated with poorer mediolateral (ML) balance. Therefore, we developed a simplified version of the ML balance assessment (MELBA) using a Microsoft-Kinect sensor, which involves tracking predictable and unpredictable targets using center-of-mass (CoM) feedback. The aim of this study was to assess and compare the impact of MS and ST on balance performance using MELBA.          

Methods: 10 ST (62±8y/o), 10 MS (36±5y/o) and 10 healthy young adults (HY, 35±3y/o) performed 2 MELBA tasks, 1 predictable (115s) and 1 unpredictable (114s). Target signals were constructed as in previous studies; however, maximum tracking frequency was set at 1.6Hz [1]. A Microsoft-Kinect sensor (30Hz) and D-flow software v3.24 (Motekforce Link, Amsterdam, The Netherlands) were used to calculate whole-body CoM. Target and CoM ML displacements were displayed on a 60” screen placed 2.5m in front of the subject. Maximum CoM displacement was set at 5.5% of the participant’s height. Balance performance was described by the phase-shift (PS) and gain (G) of the linear constant coefficient transfer function between CoM and the target signal using a Welch algorithm. Four balance descriptors per target were used to assess ML balance: f_PS and f_G, which are the frequencies at which performance drops below 90° and 0.5 for PS and G, respectively, and PS_mean and G_mean, which are the means within the bandwidth determined by f_PS and f_G. Data analysis was performed in Matlab R2017a and IBM-SPSS (Statistics 24) was used for statistical analysis.

Results: A multivariate analysis (α=0.5) performed for the four descriptors (f_PS, f_G, PS_mean and G_mean) and for the 2 targets, separately, showed significant between-group differences only for f_G for the predictable target. For this measure (f_G), ST showed the lowest value (.56Hz) with values for MS of .73Hz and .81Hz for HY. 

Discussion: These results show that neurological conditions affect the bandwidth of accurate amplitude tracking when using MELBA. This may be reflection of the impairment of physical capabilities that affect ML balance control. Although not significant, f_PS exhibited a trend similar to f_G, which may also indicate slower balance responses in ST and MS. The use of MELBA with the Kinect sensor also demonstrates the potential use of low-cost technology for objective clinical measures of balance.

Acknowledgements

ECR2017-502314 University of Melbourne

References

1. Cofré Lizama, L.E., et al., (2014). PlosOne, 9(10) e110757

Introduction: Cervical spondylotic myelopathy (CSM) is one of the most common degenerative disorders in elderly population. Although after the management of decompression surgery, patients may still demonstrate symptoms of impaired proprioception, muscle weakness, and poor motor control ability. The remained symptoms might lead to poor balance performance, clumsy gait, poor functional performance, and the following increased fall risk. Thus, these patients might benefit from the perturbation balance training, which has been suggested to be effective for improving the balance performance for the elderly. The purpose of this study is to investigate the effects of perturbation balance training using a custom-made treadmill on gait and the functional performance for patients with CSM after decompression surgery.

 

Methods: Participants aged between 50 to 80 years old, diagnosed with CSM, and underwent the decompression surgery were recruited into the study after six months of the surgery. All participants received the four-week perturbation balance training using a custom-made treadmill once per week. The baseline and follow-up assessments were performed within one week before and after the training. A three-dimensional motion analysis system (VICON) were used to collect kinematic and kinetic data. The outcomes were inclination angles (Figure. 1) in the sagittal plane between the center of mass (COM) and the center of pressure (COP) during heel-strike and toe-off in the sagittal plane, and the Timed Up and Go (TUG) Test.

 

Figure 1. Inclination angle between the center of mass (COM) and the center of pressure (COP) in the sagittal plane.

Results: Nine participants were recruited into the study (6 males and 3 females, mean age 62.45±2.46 years old). After four-week training, there were no significant differences in the inclination angles between COM and COP in the sagittal plane during heel-strike (pre-test= 11.51°±2.92°, post-test= 11.87°±2.55°; t=0.721, p=0.491) and toe-off (pre-test= 18.66°±3.36°, post-test= 18.79°±3.64°; t=0.343, p=0.741). However, the TUG Test significant improved after training (pre-test= 10.73±4.05 degrees, post-test= 8.85±2.11; t=-2.59, p=0.032).


Discussion: Although it seemed that the four-week perturbation balance treadmill training might not make difference to the gait kinematics in the sagittal plane, the functional performance of TUG Test showed significant improvement after the training. Besides, the inclination angles showed the tendency to be increased during heel-strike and toe-off. Insufficient sample size to detect the minimal detectable difference might be the reason for no significant changes in inclination angles. Therefore, further research with more subjects is needed to continue the associated investigation.

 

Acknowledgement: This study is supported by the Ministry of Science and Technology (MOST105-2628-E-002-006-MY3) and National Taiwan University (NTU105R7805) awarded to Dr. Wei-Li Hsu.

Background: Impaired postural stability is often associated with musculoskeletal or neurological pathologies, and thus can be a powerful screening and intervention target. Typical clinical methods to challenge postural stability include closing ones’ eyes and/or standing on an unstable surface. While easy to implement, these methods may not adequately detect subtle balance deficits, such as lingering effects of concussions. Recent advances in virtual reality (VR) provide a new mode to perturb postural stability [1], but most VR systems are overly complex and costly. Therefore, the purpose of this study was to demonstrate the ability of a low-cost VR stimulus to perturb postural control.

Methods: Twenty-eight healthy individuals (12/16 males/females, 25.4±8.1yrs, 1.73±0.12m; 79.1±17.5kg) performed single leg balance on firm vs. foam surfaces, and with eyes open vs. eyes closed vs. VR rollercoaster stimulus (Roller Coaster VR Attraction, Fibrum; Russia). The VR environment was created using a smart phone (LG™ V10; Huntsville, AL) and ~$10 VR headset (Google Cardboard™ V2 kit; Mountain View, CA). Center of pressure (CoP) data were recorded for 20 seconds at 1000Hz for each trial (Bertec Corp., FP4060-05-PT; Columbus, OH). 95% confidence ellipse area (EA) was calculated after low-pass Butterworth filtering CoP data at 20Hz. Additionally, sample entropy (SEn) was calculated for increment resultant CoP data after downsampling to 50Hz, without filtering. Increased EA is associated with impaired CoP control. Decreased SEn indicates more regular CoP trajectory dynamics, where less regularity is representative of healthier, more automatic postural control [2]. Wilcoxon Signed-Rank Tests were used to compare perturbation change scores (Δ= perturbation–baseline).

Results: VR had the largest destabilizing effect for both CoP measures. Spatial distribution of the CoP increased ~3X more during the VR condition (ΔEA: η=10.4cm²; p<0.001) compared to eyes closed (ΔEA: η=3.2cm²; p<0.001) and ~10X compared to foam surface (ΔEA: η=1.2cm²; p<0.001) perturbations (Fig 1A). Additionally, the VR condition had a greater effect on increasing CoP trajectory signal regularity (ΔSEn: η=-0.3; p<0.001) compared to eyes closed (ΔSEn: η=-0.2; p<0.001) and foam surface (ΔSEn: η=‑0.2; p<0.001) conditions (Fig 1B).

Discussion: A low-cost VR approach more effectively perturbed balance compared to traditional conditions of closing ones’ eyes or standing on a foam surface. The altered balance performance suggests that VR can impair spatial control of CoP (EA) as well as the automaticity of postural control (SEn) [3]. Low-cost VR appears to be a viable option to challenge postural control. Further work is needed to evaluate the efficacy of this approach for screening and rehabilitation purposes.

References:

[1] Slobounov S, et al., CyberPsych & Behavior, (2006). 9(2):188-191.

[2] Donker S, et al., Exp Brain Research, (2007). 181(1):1-11.

will complete later

Most falls in older adults result from trips, slips or pushes [1]. Assessing reactive gait stability by means of external perturbations could therefore be highly relevant to identify people at risk for falls. However, knowledge on which perturbation type discriminates best between older fallers and non-fallers is lacking. Therefore, we developed a mixed-perturbation  protocol and evaluated whether the recovery response in terms of trunk motion could distinguish between older adults with and without a history of falls.

Forty-nine older adults walked on a movable treadmill (GRAIL, Motek, The Netherlands). Trunk and lower limb kinematics were collected during steady state gait and in response to four different perturbations: ipsi- and contralateral sway platform translations, and belt accelerations and decelerations. The perturbation response was quantified as the deviation in linear and angular trunk velocities relative to steady state velocities [2]. In addition, fall-history, clinical measures and steady state gait stability were assessed.

Ten older adults were classified as fallers based on self-reported fall history. No significant differences between fallers and non-fallers were found in the clinical or steady state gait measures. Nonetheless, fallers were more affected by the contralateral sway and deceleration perturbations compared to non-fallers as indicated by significantly larger deviation in trunk motion. Reactive gait assessment thus appeared more sensitive in discriminating between older adults with and without a history of falls than conventional measures. We recommend it should therefore be used in further research seeking to identify older adults at risk for falls.

[1] Robinovitch, S.N. et al. (2013). Lancet, 381(9860): 47-54.

[2] Bruijn, S.M. et al. (2010). The Journal of Experimental Biology, 213: 3956-3952.

In athletic testing and sports medicine, balance is conventionally evaluated using a single-leg stance (SLS) test, which yields valid and reliable outcomes. However, the relatively low difficulty of this test might limit sensitivity in sports, such as soccer. Therefore, dynamic balance tests have been developed, but these tests often consist of self-induced perturbations which may vary in intensity, thereby limiting reliability of balance performance outcome measures. To improve reliability, a dynamic balance test was developed which uses standardized position controlled horizontal translational perturbations of a newly developed instrumented platform (DynSTABLE), see figure 1. This study examines the reliability of several outcome measures based on Ground Reaction Force (GRF) data in response to these perturbations. 42 (21 male and 21 female) healthy football players underwent four trials of 24 translational platform perturbations with a magnitude of 2 cm. For all evaluated outcome measures, two-way random ICC values were calculated and the correlation between measures wat determined as well. CoP velocity and a Time-to-Stabilization outcome demonstrated high ICC values (0.90 and 0.92 respectively), but only a moderate correlation (0.61). The results of this study demonstrate that position controlled platform perturbations yield reliable outcomes. CoP velocity and Time-to-Stabilization appear to measure different dimensions of stability. Possibly, these outcomes reflect utilization of an ankle and hip strategy, respectively.

Figure 1.  Subject undergoing mediolateral platform perturbations on the DynSTABLE.

Introduction

    Running or walking on a treadmill (TM), differs from over-ground locomotion; thus, the use of a TM as a replacement for over-ground locomotion has been questioned. This difference is thought to involve visual and auditory feedback (2). Visual flow refers to information obtained visually for self-motion and may provide the illusion of moving; thus it has been used as a means of adjusting locomotion (2). Visual feedback is said to be important in controlling posture during both static and dynamic situations (1). The current study examined the effect of visual flow (patterns of visual movement of surroundings) on gait attractors (3) during treadmill walking and running.

Methods

   Participants (n=16) walked (1.39 m∙s^-1) and ran (2.78 m∙s^-1) on a TM in visual flow and control conditions. A visual flow pattern of moving along a local street was projected onto three, 2.5m x 2.5m, rear projection screens located 1.5m in front and to either side of the participant. Visual flow pattern speed was chosen to elicit the illusion of walking/running with scenery moving past the participant on the sides and toward them from the front. During each walk or run three-dimensional acceleration data were collected at 300 Hz using sensors (RehaWatch, Hasomed; Magdeburg, Germany) attached to the ankles. Data collected for 3 minutes were split into three equal parts. Attractor data were determined to provide differences in individual running patterns of running (δM) and changes in running precision (δD) across speed and setting conditions (3).

Results

   Two-Way Repeated Measures ANOVA revealed a difference between speeds for δD (p < 0.05), but no difference for δM; or between visual flow and control conditions, or splits (p > 0.05) for either δM or δD. There was a significant interaction for both δM and δD (p < 0.05) (see Figure 1).

Figure 1. Mean δM (A) and δD (B) for visual flow and control conditions across speeds and time splits.

Discussion

   There was no difference in δM between visual flow or control conditions during TM walking or running across time. However, the difference for δD between speeds indicates that the attractor method can detect changes in gait variability across speeds. This seems most prevalent during the initial portion of the running as seen in Figure 1 in agreement with Weich et al. (3).

References

1. Hashiba, M. (1998) Jpn J Physiol 48: 499-504.
2. Mohler, BJ, et al. (2007) Exp Brain Res 181: 221-228.
3. Weich, C., et al. (2017) Sports Biomech, 1-14 http://www.tandfonline.com/doi/full/10.1080/14763141.2017.1391324.

Introduction: The ability to correctly shift weight is crucial for mediolateral (ML) balance. This ability is impaired in most stroke (ST) patients with neurological damage affecting the sensorimotor system and exposes them to a greater risk of falling. Due to the lack of objective tools to quantify ML balance that can be easily applicable in clinical contexts, we developed MELBA-K, a videogame based on the mediolateral balance assessment (MELBA)[1]. MELBA quantifies balance in terms of bandwidth of adequate tracking using center-of-mass (CoM) feedback on performance; however, in MELBA-K we transformed the task into a spaceship(CoM)/UFO(target) shooting videogame using a Microsoft-Kinect sensor. The aim of this study was to assess the impact of stroke on balance performance using MELBA-K.            

Methods: 10 ST (53±18y/o, 2f) and 8 healthy young adults (HY, 34±7y/o, 4f) performed 1 predictable (135s) and 1 unpredictable (132s) MELBA tasks,using the same target signals as in previous studies (0.1-2.0Hz)[1]. A Microsoft-Kinect v2 sensor and customized software were used to calculate whole-body CoM, produce target signals and display them as a spaceship and UFO, respectively. Participants were instructed to lean their whole body sideways in order to shoot, using the spaceship, at the middle of the UFO; videogame was displayed on a 60" screen placed 2.5m in front. Maximum CoM displacement was set at 5.5% of the participant’s height. Balance performance was described by the phase-shift (PS) and gain (G) of the linear constant coefficient transfer function between spaceship (CoM) and the UFO (target) using a Welch algorithm. Four balance descriptors per target were calculated: f_PS and f_G, which are the frequencies at which performance drops below 90° and 0.5 for PS and G, respectively, and PS_mean and G_mean, which are the means within the bandwidth determined by f_PS and f_G. Data analysis was performed in Matlab-R2017a and statistics on IBM-SPSS24.

Results: A multivariate analysis (α=0.5) performed for the four descriptors (f_PS, f_G, PS_mean and G_mean) and for both targets, separately, showed significant (p<0.05) between-group differences for f_PS (ST=0.45Hz, HY=0.75Hz), f_G (ST=0.45Hz, HY=0.69Hz)_, and G_mean (ST=0.70, HY=0.80) for the predictable and f_PS (ST=0.27Hz, HY=0.50Hz) and f_G (ST=0.28Hz, HY=0.67Hz) for the unpredictable target.

Discussion: These results show that neurological damage due to stroke affects the bandwidth of adequate tracking when using MELBA-K. Whereas lower PS bandwidth may indicate slow balance responses, lower G bandwidth may indicate impaired control of the musculature controlling ML displacements. Overall, MELBA-K tasks allowed quantification of sensorimotor deficits affecting ML balance control in stroke. Because MELBA-K is portable, it can be easily implemented in clinical settings.    

Acknowledgements

ECR2017-502314 University of Melbourne

References

1. Cofré Lizama, L.E., et al., (2014). PlosOne, 9(10) e110757

Introduction
Children with cerebral palsy (CP) have limited gait ability and this is considered a key target of rehabilitation. Children often walk in a flexed knee gait pattern with short step lengths. While the aetiology of gait deviations is complex, one limitation is frequently observed to be reduced ankle push-off power generation. This can be due to weakness, spasticity or disrupted motor control. This results in difficulty maintaining smooth gait progression, lacking energy transfer of centre of mass to reduce energy loss at contralateral foot strike [1]. Therapists often use verbal feedback to train gait, however, it can be difficult to convey complex biomechanical parameters in this way. Virtual reality (VR) presents a promising platform for development of intuitive and engaging biofeedback. Children with CP show the ability to adapt and improve hip and knee extension when challenged with real-time feedback during gait [2]. It is not known if they can adapt and improve more complex biomechanical parameters such as ankle power. The aim of this research was to establish the effects of real-time biofeedback in VR on clinically important gait parameters in children with cerebral palsy.
Method
Twenty-five children with spastic cerebral palsy (10y5m±2y11m, GMFCS I-II), walked on an instrumented treadmill with VR environment (GRAIL, Motek, Amsterdam). The Human Body Model [3] was used to allow for real-time calculation of biomechanical parameters from gait analysis. Following baseline gait analysis, children were challenged with patient specific targets, to improve aspects of gait with a purposeful game in which children visualised themselves as an avatar. They underwent a series of two minute trials in which they received biofeedback, while walking, challenging improved ankle power generation, knee extension and step length.
Results
During the biofeedback trials children showed the capacity to reach clinically important improvements in key aspects of gait. Peak ankle power generation was significantly increased by 36.9% (p<0.001), knee extension at initial contact increased by 7.3° (p<0.001) and step length increased 11.1% (p=0.007). While improvements were found in aspects of gait, overall gait as measured by the gait profile score was not significantly changed in any feedback trial (p=0.13).
Conclusion
These findings suggest that VR-based biofeedback is a feasible, effective method to improve clinically important aspects of gait in the short term. While individuals could improve overall gait performance, as a group, gait is not found to improve due to compensatory movements observed. Individual response may provide clinical diagnostic insight. Further investigation into the long-term training effects of biofeedback-based gait training is required to establish if this novel improved gait pattern can be refined.

[1] Kuo et al. 2010, [2] van Gelder et al. 2017, [3] van den Bogert et al. 2013

Introduction

Self-selected or prescribed walking speeds are commonly used in gait stability research. However, a prescribed walking speed will not reflect the speed at which all participants are most comfortable, stable or economical, which is problematic when comparing groups with different capacities. When using self-selected walking speeds, however, the difficulty in recovering stability following perturbations can vary with walking speed [1,2]. We therefore present a method for normalising walking speed to gait stability for the purpose of gait perturbation research.

 

Method

Ten healthy adults (24±3y) walked on the CAREN (Motekforce Link, Amsterdam) for two-minute bouts at 0.4m/s, 0.6m/s, 0.8m/s, 1.0m/s, 1.2m/s, 1.4m/s, 1.6m/s and 1.8m/s. Five retroreflective markers were attached (C7, trochanters and halluces) and were tracked by the motion capture system (100Hz; Vicon Motion Systems, Oxford, UK). The anteroposterior margins of stability (MoS) [3] adapted for a reduced kinematic model [4] were calculated at foot touchdown. To determine the stability-normalised speed, the mean MoS of the final 10 steps of each bout were fitted with a second order polynomial function. The speed that would result in MoS of 0.05m was calculated for each participant. Participants then walked for four minutes at this speed and the final 10 steps were analysed.

 

Results

A one-way repeated-measures ANOVA revealed a significant walking speed effect on MoS (P<0.0001; Fig. 1A), with differences at each speed compared to all other speeds (P<0.001). The normalisation resulted in MoS very close to the desired 0.05m (Fig. 1B) and the group-level standard deviation of the mean MoS was 0.007, lower than the prescribed speeds (range: 0.017-0.032). No significant correlations between the stability-normalised walking speeds and the participants’ height, leg length or walking speeds for a Froude number of 0.25 were found (0.39<P<0.6).

Fig. 1: Individual MoS over different walking speeds (A) and means and SDs of MoS for 10 steps at the stability-normalised walking speed (B) for each participant. The desired MoS (0.05m) is indicated by the dashed line.

 

Discussion

For a prescribed speed, even within these young healthy participants, large ranges in MoS were observed and walking speed significantly affected MoS, confirming some issues related to walking speed choice. The normalisation successfully reduced between-participant variability in MoS, meaning that the method could be beneficial for gait stability studies comparing groups with different locomotor capacities. Importantly, body dimensions were not associated with the stability-normalised walking speed, indicating that a normalisation based on body dimensions alone does not assume equivalent gait stability.

 

References

1. Bhatt et al. Gait Posture 2005;21:146-56.

2. Espy et al. Gait Posture 2010;32:378-82.

3. Hof et al. J Biomech 2005;38(1):1-8.

4. Süptitz et al. Hum Mov Sci 2013;32(6):1404-14.

INTRODUCTION

Successfully maintaining balance during upright stance is one of the major benchmarks of rehabilitation following injury to the central nervous system¹. The challenges consist in compensating or integrating the altered visual, vestibular, proprioceptive or tactile stimuli to create a dynamic equilibrium². Applying noisy stochastic stimulation increases the baseline level of these systems, resulting in lower threshold to engage afferent signaling³. In this study we investigated the effects of increased afferent signaling on balance regulation during stance in challenging environments.

METHODS

Young healthy participants were tracked using optical motion capture (100Hz, Vicon, UK) during hip-wide stance on a treadmill (GRAIL, Motek Medical) with a visual target. Following familiarization, following conditions were applied in randomized order: quiet standing, head rotations (3s pause, ω=28°/s, feedback provided), platform perturbations (δ=1s, amplitude: 50mm) in medio-lateral (ML) and antero-posterior (AP) directions. Participants completed these conditions with no stimulation (BSL) and with 90% sensory threshold noisy stochastic stimulation (0-250Hz pink noise spectrum) of the vestibular system (GVS, NeuroConn) and the foot soles (vSR, Engineering Acoustics, US). The area of the 95% confidence ellipse (A) of the center of mass (CoM) was compared between stimulations using 1-way ANOVA.

RESULTS

Quiet standing demonstrated a small area (A_CoM=46mm²), while head rotations entailed the greatest area (A_CoM=1897mm²). Stimulation modalities did not show a main effect when pooled over all conditions (F=0.15, p=0.86), however promising trends emerged. During quiet standing, both stimulation modalities resulted in increased A_CoM (GVS:+41%, vSR:+69%) (Fig.1). During head rotations, stimulation resulted in decreased A_CoM (GVS:-35%, vSR:-60%). During platform perturbations, results of A_CoM were inconclusive (ML:+10%, AP:-10%), however CoM range in the main movement direction was influenced by stimulation mode (GVS_ML:-6%, vSR_ML:-10%; GVS_AP:-20%, vSR_AP:-12%).

DISCUSSION

Selectively modulating afferent systems is a potential avenue for exploring the integration of afferent signals following neurological injury. As a first step, this initial data indicates that young healthy subjects respond to noisy vestibular and tactile stimulation. This response is increased in challenging conditions; the current data however is not sufficient to distinguish the different stimulation modi. The regulation of the center of mass seems to be a promising outcome, however a more fine-grained approach is necessary for further analysis.

FIGURE.1


REFERENCES

1.          Bohannon, R.W., and Leary, K.M. (1995). Standing balance and function over the course of acute rehabilitation. Arch. Phys. Med. Rehabil. 76, 994–996.

2.          Winter, D.A. (1995). Human blance and posture control during standing and walking. Gait Posture 3, 193–214.

3.          Moss, F., Ward, L.M., and Sannita, W.G. (2004). Stochastic resonance and sensory information processing: A tutorial and review of application. Clin. Neurophysiol. 115, 267–281.