Immediate augmented real-time forefoot weight bearing using visual feedback improves gait symmetry in chronic stroke

Abstract

BACKGROUND:

Symmetry of gait is an important component of rehabilitation in stroke patients. Insufficient weight-bearing causes gait asymmetry.

OBJECTIVE:

This study aimed to identify the immediate effects of sufficient weight-bearing on the forefoot during the stance phase using visual feedback.

METHODS:

Twenty-seven individuals with stroke enrolled in this study. All patients were evaluated for gait parameters with and without visual feedback. Visual feedback was provided through a smart application and a beam projector screen that showed a weight shift as a change in color. Spatiotemporal gait parameters were evaluated, resulting in values for a calculated symmetry index, in addition to heel % and toe % temporal values.

RESULTS:

Velocity and cadence were significantly decreased when visual feedback was provided ( $p<$ 0.05). Spatiotemporal parameters, except for bilateral step length, swing time of affected side, and single-limb support of less affected side, showed significant improvement ( $p<$ 0.05). The gait pattern of subjects was more symmetrical with visual feedback compared to non-visual feedback ( $p<$ 0.05). The toe-on time significantly improved on the affected side with visual feedback ( $p<$ 0.05).

CONCLUSION:

This study suggests that visual feedback aids in the improvement of gait symmetry, forefoot weight-bearing on the affected side, and spatiotemporal parameters.

Keywords

Biofeedback gait weight-bearing stroke

1. Introduction

Hemiplegic gait in stroke patients is characterized by reduced symmetry, reduced step length, reduced stance time on the affected side, reduced range of motion in swing phase, and reduced balance [1]. Symmetrical and functional gait can improve gait quality in stroke patients. Muscle weakness and reduced sensation on the affected side causes asymmetrical gait due to insufficient weight-bearing during stance phase [2]. Asymmetrical gait results in musculoskeletal problems on the less affected side while it causes decreased bone density [3], decreased gait speed [4], and inefficient energy consumption on the affected side [5]. Furthermore, stroke patients exhibit limited weight-shifting from the heel towards the toe during the stance phase on the paralyzed side and anterior-posterior force on the affected side is decreased at the push-off phase [6]. Insufficient weight shifting to the forefoot on the affected side causes asymmetrical gait, which progresses to decreased function. Instability due to loss of sensation, fear of falling and low balance confidence are the main contributors to decreased function [7].

In order to improve gait ability in stroke patients, visual [8], auditory [9], electromyographic [10], and vibrotactile [11] feedback have been used. A study by Walker et al. [12] reported that visual feedback could potentially prevent falls in stroke patients by providing important information on the walking environment, which leads to improved dynamic balance regulation. Furthermore, visual perception could be improved by practicing various tasks including directional discrimination [13]. A systematic review [14] showed that biofeedback has more positive effects than traditional/placebo therapy on improvement of lower extremity activities such as sitting, sit to stand, standing and walking. Moreover, these effects were reported to remain in effect for several months after the intervention. Feedback shares information on movement errors and methods of correction, which either improves the skill level or speeds up the learning process. Hence, feedback provides fundamental information to maintain motivation during learning [15].

Moreover, an observational study by Talvitie [16] reported that visual and verbal cues are the most effective means of enabling patients to understand their movements and objectives. Byl et al. [17] performed gait training in chronic stroke and Parkinson’s disease patients using visual kinematic biofeedback, and reported that visual feedback was similar to verbal feedback from a therapist. The positive effects were significantly greater than in a control group and included improved gait speed, step length, endurance, gait quality, balance, range of motion, and strength [17]. In a study of nine healthy elderly people and one stroke patient, visual feedback for toe clearance during pre-swing phase effectively reduced the risk of tripping [18]. However, most visual feedback was through center of pressure (COP) movement during gait [19] or provided graphs of kinetic data, such as timing, location, and amplitude of ground reaction forces [8, 17].

Commonly used devices that display weight distribution on the soles as colors are mostly fixed to the ground and the data are frequently provided as feedback for maintaining balance rather than for gait. When the devices are fixed to the ground, the steps are restricted to specific locations during gait. Therefore, these can cause unnatural gait and make it difficult to continuously provide visual feedback during ambulation. To regulate head and trunk movement during gait, stability is required [20]. Vision plays an important role in regulating postural stability [21]. As stability of the head and trunk decreases after stroke, the quality of visual information decreases, resulting in decreased balance [22]. Due to fear of falling and instability, most stroke patients look downwards when walking for visual compensation. Therefore, this study provided real-time visual feedback in front of the patient at eye level to improve visual stability during gait. With weight-shifting during gait, the pressure of the affected lower extremity was displayed in different colors based on intensity to emphasize anterior shifting of the COP during stance phase. Accordingly, the changes in weight-shifting during gait were provided as real-time visual feedback and the effects on spatiotemporal variables of ambulation and symmetry were investigated. The hypothesis of this study is that gait training using visual feedback will increase the forward movement of body weight on the forefoot and increase the symmetry of gait.

2. Methods

2.1 Participants

For this study, 72 stroke patients at the “H” rehabilitation hospital in Jungnang-gu, Seoul, whose symptoms occurred within the two prior years using brochures and oral public activities promotion were recruited. Selection criteria included stroke symptoms for at least six months, mini-mental status examination-Korea (MMSE-K) score of $\geqslant$ 24 [23], ability to ambulate at $>$ 0.4 m/s [24], no visual problems, and corrected eyesight of $\geqslant$ 0.7 [25]. Patients with musculoskeletal impairments, psychological disorders, or untreated medical conditions were excluded. After applying the exclusion criteria, the participants were narrowed to 10 who met all the requirements (five males, five females). All participants were given details about the study purpose and content and signed consent was obtained. The study was conducted from July 18, 2017 to July 22, 2017. Independent walking was defined as the ability to walk without supervision but could involve the use of an aid, such as a cane. Stroke was confirmed with computed tomography or magnetic resonance imaging. The study protocol has approved by the institutional review board of Sahmyook University (2-1040781-ABN-01-2017057HR).

2.2 Procedures

Figure 1.

Training with visual feedback.

This was a cross sectional study. One participant withdrew for personal reasons during the study. The remaining nine participants were asked to walk on a GAITRite ${}^{\text{TM}}$ system (GAITRite, CIR System Inc., Clifton, NJ, USA) plantar pressure measurement mat. In consideration of the natural gait measurement and acceleration and deceleration, we instructed the participants to walk two meters before the pressure measuring mat and two meters from the end of the mat [26]. The participants first walked naturally on a flat surface without visual feedback, followed by three minutes of rest and subsequent walking with visual feedback for weight shifting displayed on the screen. Using a beam projector and 80-inch screen (1,585 mm $\times$ 1,190 mm) in front of the mat, real-time visual feedback was provided (Fig. 1). The participants performed three trials of flat surface walking before and after visual feedback with a three minutes break between the trials. The patients were verbally instructed to bear sufficient weight on the affected forefoot during flat surface walking before and after visual feedback. In order to detect weight-bearing and provide visual feedback, the participants wore smart socks made of conductive material that caused a change in color on the screen with a change in weight-bearing. We used smart socks because we considered the characteristics of Koreans who take off their shoes indoors. The color was set to change from light blue to red and yellow with increasing pressure (color hex codes #93d7ee to #ffcg03) (Fig. 2). The weight-bearing information detected by the smart socks was transferred to a mobile application (Sensoria Fitness, Sensoria Inc., USA) through a Bluetooth device on the ankles.

Figure 2.

Screen showing the weight of the left forefoot during gait.

2.3 Outcome measurements

2.3.1 Temporal and spatial gait parameters

This study used a GAITRite ${}^{\text{TM}}$ system to measure temporal and spatial gait parameters through an electronic walkway connected to the USB port of a computer.

2.3.2 Heel-on time and toe-on time values

Heel-on time and toe-on time values expressed heel only and toe only [27]. Heel only (heel-on time) was defined as the time when only the heel was on the floor. Toe only (toe-on time) was defined as the time when only the toe was on the floor. These values represented two-limb support through a gait cycle. The calculation method is as follows Eqs (1) and (2). Mid on and mid off mean that the middle foot is on and off.

$\displaystyle\text{Heel only}=\text{mid on}-\text{heel on}$ (1) $\displaystyle\text{Toe only}=\text{toe off}-\text{mid off}$ (2)

2.3.3 Symmetry index

The symmetry index (SI) was calculated using the following Eq. (3), where X is the stance phase time (s) of the gait cycle. As it approaches zero, the SI reflects increasing symmetry.

$\displaystyle\text{SI}=\frac{\text{X paretic}-\text{X non}-\text{paretic}}{0.5% \left({\text{X paretic}+\text{X non}-\text{paretic}}\right)}\times 100\text{\%}$ (3)

2.4 Data analysis

Data were analyzed using SPSS version 19.0 software for Windows. The Mann-Whitney U test was used for analysis of differences in general characteristics including age, height, weight, MMSE-K score, and post-stroke duration in months. The Wilcoxon signed-rank test was used for nonparametric data since the sample size in this study was small and the data were not normally distributed by the Shapiro-Wilk test. The Wilcoxon signed-rank test was performed to identify differences within paired scores based on pre-test and post-test results for each condition. For all tests, statistical significance was set as 0.05.

3. Results

The characteristics of the nine participants are listed in Table 1.

Table 1
General characteristics of the participants

	Sex	Age (years)	Weight (kg)	Height (cm)	After stroke (months)	Stroke type	Hemi side	MAS (0/1/1 $+$ )	MMSE-K (score)
Patient 1	Female	55	72	158	15	Infarction	Right	0	21
Patient 2	Female	42	67	164	17	Infarction	Left	0	26
Patient 3	Female	45	51	168	20	Hemorrhage	Left	0	27
Patient 4	Male	58	63	159	12	Hemorrhage	Right	1	29
Patient 5	Female	45	53	157	19	Infarction	Left	0	30
Patient 6	Male	54	69	163	21	Infarction	Left	0	25
Patient 7	Female	46	55	164	15	Infarction	Right	0	28
Patient 8	Male	65	49	154	14	Hemorrhage	Right	1	30
Patient 9	Male	46	60	170	13	Infarction	Left	0	30
Median (IQR)	4/5	46 (11.5)	60 (16)	163 (9)	15 (6)	6/3	4/5	(7/2/0)	28 (4.5)

MAS: Modified Ashworth Scale. MMSE-K: mini-mental status examination-Korea.

3.1 Comparison of spatiotemporal parameters

The gait speed and cadence were significantly reduced when real-time visual feedback was provided ( $p<$ 0.05). Real-time visual feedback walking led to a significant decrease in bilateral and affected side stride length ( $p<$ 0.05), and the stance time, single-limb support, and double-limb support on the affected side were significantly improved ( $p<$ 0.05). However, bilateral step length and swing time on the affected side showed no significant differences. On the less affected side, real-time visual feedback walking led to significantly decreased stride length ( $p<$ 0.05) and significantly increased swing time, stance time, and double-limb support time ( $p<$ 0.05) (Table 2).

Table 2
Change in spatiotemporal variables with real-time visual feedback for gait

Variable	Visual feedback		Z	$p$
	Without visual feedback ( $n=$ 9)	With visual feedback ( $n=$ 9)
	Median (IQR)	Median (IQR)
Velocity (cm/s)	28.90 (29.95)	19.30 (25.35)	$-$ 2.666	0.008 ${}^{\ast}$
Cadence (steps/min)	60.40 (37.45)	40.40 (38.3)	$-$ 2.521	0.012 ${}^{\ast}$
Affected side
Step length (cm)	30.94 (11.87)	31.00 (9.04)	$-$ 1.955	0.051
Stride length (cm)	57.84 (20.82)	56.61 (16.91)	$-$ 2.666	0.008 ${}^{\ast}$
Swing time (s)	0.64 (0.21)	0.66 (0.22)	$-$ 0.533	0.594
Stance time (s)	1.36 (1.27)	2.12 (1.94)	$-$ 2.666	0.008 ${}^{\ast}$
SLS ${}^{\text{a}}$ (s)	0.40 (0.12)	0.49 (0.15)	$-$ 2.134	0.033 ${}^{\ast}$
DLS ${}^{\text{b}}$ (s)	1.01 (1.2)	1.65 (1.94)	$-$ 2.666	0.008 ${}^{\ast}$
Less affected side
Step length (cm)	27.09 (13.9)	26.03 (9.5)	$-$ 1.718	0.086
Stride length (cm)	57.71 (20.61)	56.68 (17.61)	$-$ 2.429	0.015 ${}^{\ast}$
Swing time (s)	0.40 (0.12)	0.49 (0.15)	$-$ 2.134	0.033 ${}^{\ast}$
Stance time (s)	1.61 (1.4)	2.52 (2.2)	$-$ 2.547	0.011 ${}^{\ast}$
SLS (s)	0.64 (0.21)	0.66 (0.22)	$-$ 0.533	0.594
DLS (s)	1.01 (1.25)	1.64 (1.97)	$-$ 2.666	0.008 ${}^{\ast}$

${}^{\text{a}}$ SLS: single-limb support, ${}^{\text{b}}$ DLS: double-limb support. ${}^{\ast}p<$ 0.05 by Wilcoxon signed-rank test for comparison of improvement with and without visual feedback.

Table 3

Change in symmetry index with real-time visual feedback for gait

Variable	Visual feedback		Z	$p$
	Without visual feedback ( $n=$ 9)	With visual feedback ( $n=$ 9)
	Median (IQR)	Median (IQR)
Symmetry index (%)	0.17 (0.15)	0.09 (0.10)	$-$ 2.666	0.008 ${}^{\ast}$

${}^{\ast}p<$ 0.05 by Wilcoxon signed-rank test for comparison of improvement with and without visual feedback.

Table 4

Change in heel and toe % temporal values

Variable	Visual feedback		Z	$p$
	Without visual feedback ( $n=$ 9)	With visual feedback ( $n=$ 9)
	Median (IQR)	Median (IQR)
Affected side
Toe only (%)	0.10 (0.09)	0.17 (0.11)	$-$ 2.310	0.021 ${}^{\ast}$
Heel only (%)	0.03 (0.05)	0.03 (0.10)	$-$ 0.652	0.515
Less-affected side
Toe only (%)	0.12 (0.11)	0.14 (0.17)	$-$ 1.481	0.139
Heel only (%)	0.04 (0.05)	0.05 (0.07)	$-$ 1.362	0.173

${}^{\ast}p<$ 0.05 by Wilcoxon signed-rank test for comparison of improvement with and without visual feedback.

3.2 Symmetry index, heel %, and toe % temporal values

The SI was significantly increased when real-time visual feedback was provided ( $p<$ 0.05) (Table 3). Heel % and toe % temporal values were only significantly increased for the toe on the affected side ( $p<$ 0.05) (Table 4).

4. Discussion

In this study, real-time visual feedback was performed using smart socks, and gait symmetry and spatiotemporal variables were improved by focusing on weight shifting on the affected forefoot. Gait analysis was performed based on the time of contact between the heel and the floor surface, and between the forefoot and the floor surface using spatiotemporal variables. Unlike graphing of bilateral step length using two bars as proposed in a previous study [28], or determination of COP movement using measured pressure on each plane (frontal, sagittal, transverse plane) on a Wii Fit board [29], the real-time visual feedback method in this study used changes in color based on the magnitude of weight shifting and weight bearing on the soles. The participants were intrigued by the feedback during ambulation. Kim and Krebs [30] reported that real-time visual feedback motivates participation in walking training and may induce gait adaptation, hence providing implicit motor learning. Yang et al. [29] used center of posture for visual feedback in stroke patients and reported decreased severity of pusher syndrome and improved lower extremity motor regulation. Real-time posture feedback improved concentration and promoted awareness of posture and subsequent self-adjustment [29]. The real-time visual feedback used in this study also induced immediate regulation of weight-bearing and improved the quality of gait.

In this study stance phase time was used to investigate gait symmetry because real-time visual feedback focused on weight shifting on the affected foot. Previous studies on gait symmetry reported significant improvement in the step length symmetry when swing resistance was applied using robots on the affected lower limb in stroke patients [31]. The intent was to increase muscle strength in the lower extremity to overcome resistance; however, the real-time feedback in the present study can be applied without resistance as long as visual ability is not restricted. A case study by Lewek et al. [32] also reported that visual feedback effectively improved gait symmetry. Symmetrical changes in gait can be explained as improved ability to correct gait when an error occurs during training [32].

Khallaf et al. [33] used E-med ${}^{\@setsize{\scriptsize}{9.5pt}{\viiipt}{\@viiipt}{\textregistered}}$ pedograph as visual feedback to correct equinovarus gait in stroke patients. After feedback, the center of plantar pressure significantly transitioned from the hindfoot to the forefoot. In addition to the big toe, the fourth and fifth toes have shown increased weight-bearing. In this study, the pedograph was applied in the form of smart socks to show real-time changes in weight shifting. Therefore, the participants actively shifted weight from the heel to the forefoot. When the participants walked with smart socks on, forefoot contact on the floor surface significantly increased ( $p<$ 0.05). Most stroke patients fear fully shifting their weight on the forefoot. Therefore, most stroke patients look downwards to visually confirm the location of their feet while walking. Although the weight shifts from the hindfoot to the forefoot, the patients cannot be confident as they and the therapists cannot objectively know whether the weight has fully shifted. However, the smart socks used in this study visually provide information on weight shifting and weight bearing. Byl et al. [17] used dynamic visual feedback in stroke and Parkinson’s disease patients, and reported that visual feedback is similar to verbal feedback from a therapist, resulting in significantly improve gait speed, step length, endurance, balance, range of motion, and strength. Hamacher et al. [34] reported that speed decreases as focus on gait precision increases. In this study, the decrease in gait speed and cadence could also be explained by efforts to accurately perform motion based on real-time visual feedback. Visual feedback walking could improve speed as well as weight support and walking symmetry if sufficient time is allowed for subjects to perform voluntary walking.

In this study, real-time visual feedback was provided using a large frontal screen for gait. The advantage of this type of intervention is that stroke patients can perform gait by fixing their gaze to the front rather than looking at the feet or walking with a smartphone application based on walking. However, when using a large screen during walking, it is difficult to provide continuous visual feedback when changing directions. Limitations of this study include the small number of participants, which impedes generalization of the results. This study was limited by space constraints in the ability to provide real-time visual feedback. In addition, different sizes of socks were needed to accommodate different foot sizes. Although the stroke participants in this study (mean age of 50.67 years) reported that the visual color display was very good, objective satisfaction was not studied. In further studies, the participant satisfaction should be confirmed using objective indicators and changes in other gait variables other than symmetry should be investigated for a longer period.

5. Conclusion

In conclusion, real-time visual feedback was used to induce weight-bearing on the forefoot, and significantly improved gait symmetry and quality in stroke patients. In addition to quantitative changes, qualitative improvement in gait is very important for satisfaction in stroke patients, with the goal of early discharge and return to society.

Footnotes

Acknowledgments

This study was supported by Sahmyook University.

Conflict of interest

None to report.

References

Druzbicki

Guzik

Przysada

Kwolek

Brzozowska-Magon

Sobolewski

. Changes in gait symmetry after training on a treadmill with biofeedback in chronic stroke patients: A 6-month follow-up from a randomized controlled trial. Med Sci Monit. 2016; 22: 4859-68.

Krishnan

Khoo

Marayong

DeMars

Cormack

. Gait training in chronic stroke using walk-even feedback device: A pilot study. Neuroscience Journal. 2016; 2016: 6808319.

Jorgensen

Crabtree

Reeve

Jacobsen

. Ambulatory level and asymmetrical weight bearing after stroke affects bone loss in the upper and lower part of the femoral neck differently: Bone adaptation after decreased mechanical loading. Bone. 2000; 27(5): 701-7.

Patterson

Parafianowicz

Danells

Closson

Verrier

Staines

, et al. Gait asymmetry in community-ambulating stroke survivors. Archives of Physical Medicine and Rehabilitation. 2008; 89(2): 304-10.

Platts

Rafferty

Paul

. Metabolic cost of over ground gait in younger stroke patients and healthy controls. Medicine and Science in Sports and Exercise. 2006; 38(6): 1041-6.

Tokuno

Eng

. Gait initiation is dependent on the function of the paretic trailing limb in individuals with stroke. Gait Posture. 2006; 24(4): 424-8.

Schinkel-Ivy

Inness

Mansfield

. Relationships between fear of falling, balance confidence, and control of balance, gait, and reactive stepping in individuals with sub-acute stroke. Gait Posture. 2016; 43: 154-9.

Massaad

Lejeune

Detrembleur

. Reducing the energy cost of hemiparetic gait using center of mass feedback: A pilot study. Neurorehabilitation and Neural Repair. 2010; 24(4): 338-47.

Choi

Lee

. Effect of the cognitive-motor dual-task using auditory cue on balance of surviviors with chronic stroke: A pilot study. Clinical Rehabilitation. 2015; 29(8): 763-70.

10.

Jonsdottir

Cattaneo

Regola

Crippa

Recalcati

Rabuffetti

, et al. Concepts of motor learning applied to a rehabilitation protocol using biofeedback to improve gait in a chronic stroke patient: An A-B system study with multiple gait analyses. Neurorehabilitation and Neural Repair. 2007; 21(2): 190-4.

11.

Afzal

Lee

Park

Yoon

. A portable gait asymmetry rehabilitation system for individuals with stroke using a vibrotactile feedback. Biomed Res Int. 2015; 2015: 375638.

12.

Walker

Hyngstrom

Schmit

. Influence of visual feedback on dynamic balance control in chronic stroke survivors. Journal of Biomechanics. 2016; 49(5): 698-703.

13.

Funk

Finke

Reinhart

Kardinal

Utz

Rosenthal

, et al. Effects of feedback-based visual line-orientation discrimination training for visuospatial disorders after stroke. Neurorehabilitation and Neural Repair. 2013; 27(2): 142-52.

14.

Stanton

Ada

Dean

Preston

. Biofeedback improves activities of the lower limb after stroke: A systematic review. Journal of Physiotherapy. 2011; 57(3): 145-55.

15.

Van Vliet

Wulf

. Extrinsic feedback for motor learning after stroke: What is the evidence? Disability and Rehabilitation. 2006; 28(13-14): 831-40.

16.

Talvitie

. Socio-affective characteristics and properties of extrinsic feedback in physiotherapy. Physiotherapy Research International: The Journal for Researchers and Clinicians in Physical Therapy. 2000; 5(3): 173-89.

17.

Byl

Zhang

Coo

Tomizuka

. Clinical impact of gait training enhanced with visual kinematic biofeedback: Patients with Parkinson’s disease and patients stable post stroke. Neuropsychologia. 2015; 79(Pt B): 332-43.

18.

Begg

Tirosh

Said

Sparrow

Steinberg

Levinger

, et al. Gait training with real-time augmented toe-ground clearance information decreases tripping risk in older adults and a person with chronic stroke. Frontiers in Human Neuroscience. 2014; 8: 243.

19.

Kim

. The effects of symmetric center of pressure displacement training with feedback on the gait of stroke patients. Journal of Physical Therapy Science. 2015; 27(3): 855-7.

20.

Pimenta

Correia

Alves

Virella

. Effects of oculomotor and gaze stability exercises on balance after stroke: Clinical trial protocol. Porto Biomedical Journal. 2017.

21.

Aoki

Otani

Morishita

Domen

. The effects of various visual conditions on trunk control during ambulation in chronic post stroke patients. Gait Posture. 2017; 52: 301-7.

22.

Lamontagne

De Serres

Fung

Paquet

. Stroke affects the coordination and stabilization of head, thorax and pelvis during voluntary horizontal head motions performed in walking. Clinical Neurophysiology. 2005; 116(1): 101-11.

23.

Kim

Choi

Lee

Song

. Virtual dual-task treadmill training using video recording for gait of chronic stroke survivors: A randomized controlled trial. Journal of Physical Therapy Science. 2015; 27(12): 3693-7.

24.

Dean

Ada

Lindley

. Treadmill training provides greater benefit to the subgroup of community-dwelling people after stroke who walk faster than 04 m/s: A randomised trial. Journal of Physiotherapy. 2014; 60(2): 97-101.

25.

Young Hoon

Joo Hwa

. Correlation between disc size and retinal nerve fiber layer thickness in normal tension glaucoma. Jounal of The Korean Ophtalmological Societry. 2008; 49(5): 784-90.

26.

Donoghue

Savva

Cronin

Kenny

Horgan

. Using timed up and go and usual gait speed to predict incident disability in daily activities among community-dwelling adults aged 65 and older. Archives of Physical Medicine and Rehabilitation. 2014; 95(10): 1954-61.

27.

Ade

Schalkwijk

Psarakis

Laporte

Faras

Sandoval

, et al. Between session reliability of heel-to-toe progression measurements in the stance phase of gait. PloS One. 2018; 13(7): e0200436.

28.

Kim

Kayitesi

Chan

Graham

. Effects of partial absence of visual feedback information on gait symmetry. Appl Psychophysiol Biofeedback. 2017; 42(2): 107-15.

29.

Yang

Y-R

Chen

Y-H

Chang

H-C

Chan

R-C

Wei

S-H

Wang

R-Y

. Effects of interactive visual feedback training on post-stroke pusher syndrome: A pilot randomized controlled study. Clinical Rehabilitation. 2015; 29(10): 987-93.

30.

Kim

Krebs

. Effects of implicit visual feedback distortion on human gait. Experimental Brain Research. 2012; 218(3): 495-502.

31.

Yen

Schmit

. Using swing resistance and assistance to improve gait symmetry in individuals post-stroke. Hum Mov Sci. 2015; 42: 212-24.

32.

Lewek

Feasel

Wentz

Brooks

, Jr. Whitton

. Use of visual and proprioceptive feedback to improve gait speed and spatiotemporal symmetry following chronic stroke: A case series. Physical Therapy. 2012; 92(5): 748-56.

33.

Khallaf

Gabr

Fayed

. Effect of task specific exercises, gait training, and visual biofeedback on equinovarus gait among individuals with stroke: Randomized controlled study. Neurology Research International. 2014; 2014: 693048.

34.

Hamacher

Bertram

Fölsch

Schega

. Evaluation of a visual feedback system in gait retraining: A pilot study. Gait Posture. 2012; 36(2): 182-6.