Human Movement Science xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Human Movement Science journal homepage: www.elsevier.com/locate/humov

Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury Sheng-Che Yen a, Jill M. Landry b, Ming Wu b,c,⇑ a

Department of Physical Therapy, Northeastern University, Boston, MA 02115, USA Sensory Motor Performance Program, Rehabilitation Institute of Chicago, Chicago, IL 60611, USA c Department of PM & R, Northwestern University Medical School, Chicago, IL 60611, USA b

a r t i c l e

i n f o

Article history: Available online xxxx PsycINFO classification: 2330 2320 3380 Keywords: Locomotion Visual feedback Proprioceptive feedback Resistance Spinal cord injury

a b s t r a c t Different forms of augmented feedback may engage different motor learning pathways, but it is unclear how these pathways interact with each other, especially in patients with incomplete spinal cord injury (SCI). The purpose of this study was to test whether augmented multisensory feedback could enhance aftereffects following short term locomotor training (i.e., adaptation) in patients with incomplete SCI. A total of 10 subjects with incomplete SCI were recruited to perform locomotor adaptation. Three types of augmented feedback were provided during the adaptation: (a) computerized visual cues showing the actual and target stride length (augmented visual feedback); (b) a swing resistance applied to the leg (augmented proprioceptive feedback); (c) a combination of the visual cues and resistance (augmented multisensory feedback). The results showed that subjects’ stride length increased in all conditions following the adaptation, but the increase was greater and retained longer in the multisensory feedback condition. The multisensory feedback provided in this study may engage both explicit and implicit learning pathways during the adaptation and in turn enhance the aftereffect. The results implied that multisensory feedback may be used as an adjunctive approach to enhance gait recovery in humans with SCI. Ó 2014 Elsevier B.V. All rights reserved.

⇑ Corresponding author at: Sensory Motor Performance Program, Rehabilitation Institute of Chicago, 345 East Superior Street, Room 1406, Chicago, IL 60611, USA. Tel.: +1 312 238 0700; fax: +1 312 238 2208. E-mail address: [email protected] (M. Wu). http://dx.doi.org/10.1016/j.humov.2014.03.006 0167-9457/Ó 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

2

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

1. Introduction One of the most common goals of rehabilitation in patients with incomplete spinal cord injury (SCI) is to regain functional walking (Anderson, 2004), as limitations in mobility can adversely affect most activities of daily living (Noreau, Fougeyrollas, Post, & Asano, 2005; Post & Noreau, 2005). Afferent inputs play an important role in gait training, as they have the potential to facilitate neuroplasticity associated with improved walking following a spinal lesion (Barbeau, Fung, Leroux, & Ladouceur, 2002). Patients with an incomplete SCI usually experience impaired afferent inputs due to damage of the sensory ascending pathways. As a result, enhancing afferent inputs may be crucial in gait rehabilitation for patients with an incomplete SCI. Providing augmented feedback is one approach to enhance afferent inputs. Augmented feedback can be provided through different sensory channels. Adding a swing resistance to the leg during walking is one way to augment proprioception, as the resistance stimulates proprioceptors such as muscle spindles and Golgi tendon organ and creates error signals through the proprioceptive pathways (Lam & Pearson, 2001). Studies on patients with incomplete SCI have shown that swing resistance can enhance the output of the leg flexors and induce an aftereffect consisting of an increase in step/stride length following load release (Houldin, Luttin, & Lam, 2011; Lam, Wirz, Lünenburger, & Dietz, 2008; Yen, Schmit, Landry, Roth, & Wu, 2012). The occurrence of aftereffect has been thought of as an indicator that central command has been updated (Blanchette & Bouyer, 2009) and is a phenomenon compatible to motor learning (Fortin, Blanchette, McFadyen, & Bouyer, 2009). In fact, the force perturbation paradigm has been regarded as implicit learning as the patients learned to take longer strides without a conscious decision to do so (Patton & Mussa-Ivaldi, 2004). On the other hand, augmented visual feedback is a commonly used approach to help patients with incomplete SCI detect stepping errors. For example, clinicians usually place visual cues on the floor to help patients recognize the difference between the expected and actual stride lengths (Amatachaya, Keawsutthi, Amatachaya, & Manimmanakorn, 2009). According to the visual goal, patients modify their motor plan to minimize errors. This process requires the engagement of cognitive process (i.e., a conscious decision is made by the patient to take a longer step), and is regarded as explicit learning. While the swing resistance and augmented visual feedback may induce different motor learning processes, both of them may modulate the patients’ stride/step length through similar pathways, including: (a) modification of motor commands for stepping at the supraspinal level; (b) enhancement of neural descending drive in the residual spinal pathways. Specifically, error signals detected by the visual and proprioceptive channels can induce motor adaption, causing recalibration of motor command for stepping to minimize the different between the actual and expected stride/step length (Bastian, 2008). The neural descending drive appears to increase when one moves against resistance (Aagaard, Simonsen, Andersen, Magnusson, & Dyhre-Poulsen, 2002; Sale, 1988). Providing augmented visual feedback during gait training can enhance active involvement and thus increase descending drive and motor outputs (Banz, Bolliger, Colombo, Dietz, & Lunenburger, 2008). This leads us to postulate that providing the swing resistance and augmented feedback together (multisensory feedback) may enhance gait training outcomes compared to providing either type of feedback alone (unisensory feedback) in patients with incomplete SCI. Specifically, multisensory feedback may further augment neural descending drive in the residual spinal pathways. Also, error-driven learning may be more effective (i.e., effects can retain longer) when different learning pathways (implicit and explicit) are simultaneously engaged. However, literature is contradictory on the effect of combining implicit and explicit learning on motor training outcomes. While some investigators reported beneficial effects of explicit information on implicit motor learning (Boyd & Winstein, 2001; Curran & Keele, 1993), others reported detrimental effects (Green & Flowers, 1991; Malone & Bastian, 2010; Reber, 1976) or modest consequences (Reber & Squire, 1998; Shea, Wulf, Whitacre, & Park, 2001). These contradictory findings may result from some combined factors such as task difference, the type, timing, and salience of explicit information, and the characteristics of participants (Boyd & Winstein, 2004). In particular, we are not aware of any studies investigating the impact of explicit information on implicit motor learning in patients with SCI during locomotion. Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

3

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

The purpose of this study was to use short term locomotor adaptation as a model to generate some insights into the effects of multisensory and unisensory feedback on motor learning in patients with incomplete SCI. We used a robot-generated force perturbation (augmented proprioceptive feedback) and/or a customized visual feedback device to induce an aftereffect consisting of an increase in stride length. We hypothesized that the aftereffect would be longer in retention and greater in magnitude in the multisensory condition than that in the unisensory condition.

2. Methods 2.1. Subjects A total of 10 subjects with incomplete SCI were recruited. Their clinical features are described in Table 1. The inclusion criteria included: (1) age between 18 and 65 years, (2) level of injury between C1-T10, (3) AIS level at C or D, (4) ability to ambulate overground with assistive devices (e.g., cane or walker) as needed, or with orthotics that do not cross the knee. Exclusion criteria for this study included the presence of multiple CNS lesion sites, urinary tract infection, other secondary infections, heterotopic ossification, respiratory insufficiency, history of fracture due to osteoporosis, or the inability to give informed consent. Informed consent was obtained and all procedures were conducted in accordance with the Helsinki Declaration of 1975 and approved by the Institutional Review Boards of Northwestern University, Chicago, IL. 2.2. Instrumentations Augmented proprioceptive feedback was achieved by using a controlled load that resists leg swing during gait. A custom designed cable driven robotic system was used to provide the resistance load (Fig. 1A). A detailed description for the system has been reported previously (Wu, Hornby, Landry, Roth, & Schmit, 2011). Briefly, the system works in conjunction with a motorized treadmill and a body weight support system. It consists of four nylon-coated stainless-steel cables driven by four motors and cable spools. In this study, one of the cables was attached to the subject’s right leg at the ankle to provide resistance load during the swing phase of gait (Fig. 1B). The load was applied from the late stance phase to the mid swing phase of gait. Leg swing is initiated in the late stance phase and is decelerated in the mid swing phase (Perry, 1992). Afferents input from hip flexors have been shown to modulate the locomotor activities in humans with SCI (Dietz, Müller, & Colombo, 2002). Recent studies indicated this modulation effect is phase dependent with a stronger effect during the stance to swing transition (Wu, Gordon, Kahn, & Schmit, 2011). We postulated that providing resistance during this period is most likely to enhance the leg swing activity and result in an increase in stride length.

Table 1 Subject information. Case

Age

Years post injury

AIS level

Level of injury

BWS (% BW)

Test speed (m/s)

Load (N)

MVC (N)

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10

48 54 39 50 47 28 46 64 57 51

3 10 6 8 27 1 15 4 16 32

D D D D D D D D D D

C6–C7 C1–C6 C4 C6–C7 C5–C6 C7 C5–C7 C4 C5–C7 C4–C6

0 0 0 0 10 0 0 10 0 0

0.54 0.74 0.46 0.7 0.5 0.89 0.7 0.43 0.63 0.6

17 18 19 17 15 26 25 14 21 14

110 178 120 78 70 150 167 70 121 74

Abbreviations: AIS = American spinal injury association impairment scale; MVC = maximum voluntary isometric contraction; BWS = body weight support; BW = body weight.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

4

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

Fig. 1. (A) Illustration of the cable robot. (B) Application of resistance load. (C) Illustration of visual feedback device.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

5

A customized 3-dimensional (3D) position sensor was used to measure the ankle position during treadmill walking (Fig. 1B). The description of the sensor has been provided in detail elsewhere (Yen et al., 2012). The events of heel contact and toe-off during treadmill walking were identified from the ankle position data to segment walking cycles. Specifically, heel contact was defined as the time during which the ankle trajectory changed its moving direction from forward to backward; toe-off was defined as the time during which the ankle trajectory changed its moving direction from backward to forward (Zeni, Richards, & Higginson, 2008). A customized program written in LabVIEW language (National Instruments, Austin, TX) was used to acquire the ankle position data as well as to output the load command signals to the servomotor systems. The method for visual feedback was developed through a computerized program written in LabVIEW language, and was presented on a monitor positioned in front of the treadmill (Fig. 1A). The program consisted of a moving object, a target, and a feedback LED (Fig. 1C). The moving object reflected the subject’s ankle position in the anterior–posterior direction detected by the 3D ankle position sensor during treadmill walking. The target was a pre-defined visual goal for the subject to reach during the swing phase. The feedback LED was highlighted when the moving object exceeded the target. The location of the target was set at a position that was 10% further anterior to the subject’s typical ankle position during heel contact, as measured by the ankle position sensor. The choice of this position was based on our pilot work which demonstrated that some subjects had difficulty reaching a target location set at a position greater than 10% further anterior during treadmill walking for more than 10 min. Each subject’s typical ankle position during heel contact was determined prior to the data collection and was measured as the mean position across 20 gait cycles. To ensure that the subject maintained the same position on the treadmill throughout the data collection process, they were instructed to walk while maintaining their abdomen lightly against a fixed front handrail.

2.3. Procedures Subjects were instructed to walk on a treadmill in 3 test conditions: (a) augmented visual feedback (visual feedback); (b) augmented proprioceptive feedback (resistance); (c) augmented visual and proprioceptive feedback (combination). The order of the conditions was randomized. There was a 10-min break between the test conditions, and the subject was asked to sit down during each break. The test speed in each condition was set at each subject’s self-selected comfortable speed while walking on a treadmill (without receiving any augmented feedback). The test speed was determined prior to data collection and is presented in Table 1. Each test condition consisted of 3 periods, including baseline, adaptation, and post adaptation. In the baseline period, subjects walked on the treadmill without any augmented feedback for 1 min. In the adaptation period, subjects walked with the assigned feedback for 7 min. In the post adaptation period, we removed the feedback while subjects continued walking on the treadmill for an additional 2 min. One minute prior to entering the post-adaptation period, subjects were instructed to ‘‘walk comfortably after the resistance and/or computerized visual feedback was removed.’’ Subjects were allowed to use the front handrail as a support during walking, and they completed the baseline, adaptation, and post adaptation periods continuously without a break. To minimize the possibility that subjects use the arms to pull the trunk and in turn the leg forward to artificially increase the stride length, we instructed them to walk while maintaining their abdomen lightly against the front handrail. We verbally reminded subjects to maintain this position when they showed signs of drifting away. During the adaptation period with the visual feedback condition, subjects were instructed to shift the moving object towards the target on the monitor by taking a longer stride length. During the adaptation period with the resistance condition, subjects walked with a controlled resistance load added to the right leg at the ankle. The load was set at 20% of each subject’s maximum voluntary isometric contraction (MVC) when performing hip flexion (with full knee extension) in a standing position in order to normalize the challenge level of resistance load across subjects with SCI. The normalization was performed relative to a hip flexion MVC with full knee extension because this is the posture similar to the late stance when resistance load was applied during walking. Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

6

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

If subjects felt it was too heavy for them to walk with load for 7 min, the load amount was adjusted. The MVC was recorded using a cable and load cell (Transducer Techniques, Temecula, CA), which was attached to a rigid frame located at the back of the treadmill. Three trials were repeated for each subject with the average of the 3 trials was used to determine the load. The resistance load and MVC are reported in Table 1. During the adaptation period with the combination condition, subjects walked with the resistance load while working to shift the moving object towards the target in the computerized visual feedback device. The EMG activity of the right Tibialis Anterior (TA), Medial Gastrocnemious (MG), Soleus (SO), Vastus Medialis (VM), Rectus Femoris (RF), Medial Hamstrings (MH), and left TA (LTA) and MG (LMG), were recorded throughout the experiment. Ag–AgCl surface electrodes (ConMed, UTICA, NY) were applied with a 2.5 cm center-to-center spacing over the belly of each muscle on lightly abraded skin. Active preamplifiers with shielded leads were attached to the electrodes and connected to an Octopus AMT-8 EMG unit (Bortec Biomedical Ltd, Calgary, Alberta, Canada). All channels were amplified (gain = 500) and sampled (500 Hz) using the same computer.

2.4. Data analysis and statistical analysis The primary measure of this study was stride length of the right leg. In this study, stride length was defined as the distance between toe-off and the following heel contact of the same leg (i.e., the distance of leg swing). The swing time and step height of the right leg were also calculated to provide a more complete observation of kinematic adaptation. Step height was defined as the absolute difference between the highest and lowest points of the ankle position during a gait cycle. We also calculated the symmetry index (SI) of stride length, swing time, and step height to understand whether and how gait symmetry changed in each feedback condition. The SI was calculated as:

SI ¼

XR  XL 0:5  ðXR þ XLÞ

ð1Þ

where XR is the right leg data and XL is the left leg data (Robinson, Herzog, & Nigg, 1987). A SI score that is closer to 0 suggests a more symmetrical gait parameter. The surface EMG signals from the muscles tested were low-pass filtered at 250 Hz, high-pass filtered at 10 Hz, and notch filtered at 60 Hz using a second-order Butterworth filter before rectification. The EMG data were then integrated from the last 20% of the stance phase to the first 20% of swing phase (IEMG data). This period was selected because this was the time period when the resistance load was applied. The integrated EMGs were used to compare the changes in muscle activity across different feedback conditions. After each outcome measure was calculated for each stride, the data were divided into bins by taking the average of five consecutive strides. The baseline value was represented by the second last bin of the baseline period. We assumed that the subjects had reached a steady state of walking by this time period, and were not making any changes in their gait pattern in anticipation to the upcoming force perturbation (which was likely to occur in the last bin of the baseline period). We compared the baseline value to the 1st bin of the post adaptation period (post 1) to understand whether the feedback induced a significant aftereffect. For variables that showed a significant aftereffect, we continued to compare the baseline value to the 5th and 10th bins of the post adaptation period (post 5 and post 10) to understand whether the aftereffect lasted over time. Each comparison was done using a paired-t test. Due to multiple comparisons (baseline versus post 1 and post 5), the p values were Bonferroni adjusted. For variables that showed a significant aftereffect, we also compared the magnitude of the aftereffect during the first 5 bins of the post-adaptation period across the three test conditions. The magnitude of aftereffects was represented by the difference score between each bin in the post-adaptation period and the baseline value. The comparison was done using linear mixed model (LMM) with compound symmetry covariance structure. All statistical analyses were conducted using SPSS version 19 (SPSS, Chicago, IL). The a level for all analyses was set at 0.05. Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

7

Fig. 2. (A) The group’s mean change of stride length from the baseline during the adaptation (early and late) and postadaptation (post 1 to post 10) periods. The horizontal dashed line represent difference score = 0, i.e., no change from the baseline. The gray area represents the period with feedback application. (B) The mean stride length during the baseline, post 1, and post 10 in the combination condition. (C) The mean stride length during the baseline, post 1, and post 10 in the visual feedback condition. (D) The mean stride length during the baseline, post 1, and post 10 in the resistance condition. (E) The mean stride length during the first 5 bins of the post-adaptation period in all conditions. (F) The group’s mean change of stride length symmetry from the baseline during the adaptation and post-adaptation periods. The horizontal dashed line represent difference score = 0, i.e., no change from the baseline. The gray area represents the period with feedback application. In all figures, the error bars represent standard deviation. ⁄p < 0.05.

3. Results 3.1. Stride length All three types of augmented feedback induced an aftereffect consisting of an increase in stride length during the post adaptation period (Fig. 2A). Specifically, the stride length in post 1 was significantly Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

8

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

greater than the baseline in all three conditions (p < 0.02, Fig. 2B–D). In addition, the aftereffect lasted longer in the combination condition than those in the visual feedback and resistance conditions. For instance, in the combination condition, the stride length remained above the baseline at post 5 (p = 0.01) and post 10 (p = 0.02) (Fig. 2B). In contrast, no significant difference was found between the baseline and post 5, and between the baseline and post 10 in the visual feedback (p > 0.1, Fig. 2C) or resistance conditions (p > 0.2, Fig. 2D). The LMM detected a significant fixed effect of condition on the magnitude of the aftereffect during the first 5 bins of the post-adaptation period (p = 0.02). Post-hoc tests with Bonferroni correction indicated that the magnitude was greater in the combination condition compared to the resistance condition (p = 0.01, Fig. 2E). No significant differences were detected in the other pairs of comparison (p > 0.3). The symmetry of the stride length slightly deviated from the baseline during both adaptation and post adaptation periods in all three conditions (Fig. 2F). The stride length symmetry did not differ between the baseline and post 1 in all three conditions (p > 0.1), i.e., no aftereffect. 3.2. Swing time During the adaptation and post-adaptation periods, the swing time of the right leg tended to decrease from the baseline in the combination condition, while it tended to increase from the baseline in the visual feedback and resistance conditions (Fig. 3A). However, these changes were small in magnitude. The swing time in post 1 was not significantly different from the baseline in all conditions (p > 0.2), suggesting no aftereffect. The swing time symmetry had a slight deviation from the baseline during the adaptation and post-adaptation periods in all three conditions (Fig. 3B). Comparing the baseline and post 1 revealed that none of the augmented feedbacks induced significant aftereffects in swing time symmetry (p > 0.5).

A

B

C

D

Fig. 3. The group’s mean change of swing time (A), swing time symmetry (B), step height (C), and step height symmetry (D) from the baseline during the adaptation and post-adaptation periods. In all figures, the error bars represent standard deviation. The horizontal dashed line represent difference score = 0, i.e., no change from the baseline. The gray area represents the period with feedback application.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

9

3.3. Step height The step height of the right leg increased from the baseline during the early and late adaptation period in all conditions (Fig. 3C) but washed out during the post-adaptation period (Fig. 3C). The step height symmetry also demonstrated a similar pattern (Fig. 3D). Paired t tests indicated that the difference in both step height (p > 0.3) and step height symmetry (p > 0.4) was not significant between the baseline and post 1 in all conditions, suggesting no aftereffect.

Fig. 4. The group’s mean EMG pattern (linear enveloped) of all muscles examined during the baseline, early adaptation period, late adaptation period, post 1, and post 10 in each test condition.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

10

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

3.4. EMG Fig. 4 shows the mean EMG patterns across all subjects of all muscles examined in each test condition. From the late stance to mid-swing period (gray area in Fig. 4, when the load was applied in the resistance and combination conditions) of the right leg, we observed an increase in EMG activity in the RF during the early and late adaptation period as well as post 1 and post 10, especially in the combination condition and the visual feedback condition, although IEMG data were not significant due to large variability. Similar increases in EMG activities can be observed in LTA, and the increase was larger in the combination condition compared to the visual feedback and resistance conditions, although IEMG data were not significant due to large variability. The VM, MG, and SO EMG activities did not have much change from the late stance to mid swing period in all conditions. However, the VM showed an increase in EMG activity in the late swing period during the adaptation and post-adaptation periods, particularly in the resistance and combination conditions. The TA EMG activity did not have much change from the late stance to mid swing period during the visual feedback and resistance conditions. However, it showed a decrease in the combination condition, especially during the late adaptation period and post 1 and 10. Similarly, the MH and LMG also showed decreases in EMG activities from the late stance to mid swing period of the right leg in the combination condition. Case 7 was removed from the analysis for all right leg muscles and case 2 was removed from the analysis of LMG both due to technical problems during the data collection process. The IEMG data indicated that the RF (Fig. 5A) and LTA (Fig. 5B) had increases in EMG activity during post 1 compared to baseline in all conditions. However, none of these increases reached statistical significance (for RF, p > 0.14; for LTA, p > 0.08), suggesting that the aftereffect was not significant. We also did not detect significant aftereffect in the other six muscles (p > 0.16). 4. Discussion Our data demonstrated that the combination of visual feedback and resistance induced a longer lasting and greater aftereffect in stride length compared to the conditions with visual feedback or

Fig. 5. The IEMG data of RF (A) and LTA (B) in baseline and post 1 in all conditions.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

11

resistance alone in patients with incomplete SCI. The two unisensory feedback conditions (visual feedback and resistance) may elicit different learning processes for increasing stride length. Specifically, visual feedback led to an ‘‘explicit’’ learning process as subjects consciously increased their stride length in order to reach the pre-defined visual goal on the visual feedback device. On the other hand, the resistance load elicited motor adaptation, which was an ‘‘implicit’’ learning process (Patton & Mussa-Ivaldi, 2004). Subjects subconsciously increased their stride length through adaption to the resistance load for a period of time. Our results suggest that engaging both implicit and explicit pathways could enhance motor learning effects. Different neural networks may be involved in the various feedback conditions during the adaptation process. Visual feedback may engage the posterior parietal cortex, which is known to contribute to the planning and execution of visually guided locomotion (Beloozerova & Sirota, 2003). Whereas, the addition of resistance may engage the cerebellum, which is an essential neural network for locomotor adaptation to a proprioceptive perturbation (Morton & Bastian, 2006). Studies have indicated that the posterior parietal cortex and the cerebellum are connected (Ramnani, 2012), which may explain why providing both visual feedback and a resistance load enhanced the aftereffect in our subjects with incomplete SCI. Specifically, multisensory feedback may further augment neural descending drive in the residual spinal pathways, which may increase the excitability of the spinal cord locomotor circuitry, resulting in an improved walking performance. In addition, multisensory feedback may activate neural networks to a greater extent than unisensory feedback, and therefore induce a stronger motor memory. This is consistent with a previous imaging study showing that compared to unisensory feedback, multisensory feedback led to a greater activation of the hippocampus, which plays an important role in consolidating memory (Butler & James, 2011). The multisensory condition may further enhance subjects’ effort in leg swing compared to the unisensory conditions. The swing time tended to increase in the resistance condition as the force can slow down the leg swing velocity. The swing time also tended to increase in the visual feedback condition as subjects needed to increase the distance of leg swing (and therefore the time) in order to hit the target in the visual feedback device. However, in the combination condition, the swing time was reduced with an increase in stride length, suggesting that subjects increased effort to increase swing velocity. A possible reason for this is that subjects took a greater voluntary effort to overcome the resistance to reach the target in the combination condition. Previous results have suggested that voluntary effort may benefit motor memory formation (Kaelin-Lang, Sawaki, & Cohen, 2005). This may also explain why the aftereffect retained longer in the multisensory condition compared to unisensory conditions. In a previous study on split-belt treadmill adaptation, the retention of the aftereffect became shorter when visual instruction was given to healthy subjects (Malone & Bastian, 2010). These findings were different from what we observed in our subjects with incomplete SCI, suggesting that spinal lesions may change how individuals process visual and proprioceptive information. Specifically, visual feedback may be redundant to proprioceptive feedback in healthy individuals, but this may be not case in patients with incomplete SCI. Spinal lesions may impair patients’ ability to learn from proprioceptive feedback alone. Therefore, it becomes important to enhance feedback through additional methods (in this instance, visual feedback) that complements the proprioceptive information and facilitates the adaptation process. In addition, conscious control is usually not needed during walking in healthy individuals (Malone & Bastian, 2010). Providing visual instruction engages the conscious learning process, which may interfere with automatic control and compromise spontaneous adaptation. In patients with incomplete SCI, however, automatic control of walking is compromised and supraspinal control becomes essential. This may explain why visual feedback facilitated locomotor adaptation and strengthened the aftereffect in our subjects with incomplete SCI. Previous studies have shown that applying swing phase resistance to a leg can induce an aftereffect consisting of an increase in the RF activity in patients with incomplete SCI (Houldin et al., 2011; Lam, Anderschitz, & Dietz, 2006; Lam et al., 2008). In this current study, we also observed increases in the RF, although the increase did not reach statistical significance. Our subjects demonstrated large variability in the RF activity, as indicated by the long error bars in Fig. 5A, which may explain why the increase did not reach statistical significance. The increase in the RF activity from the late stance phase to the mid swing phase was synergized with a decrease in MH activity in the combination condition, Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

12

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

while the MH activity did not have much change in the visual feedback and resistance condition. The more coordinated change in the RF and MH activations may explain why the increase in stride length was longer in the combination condition. Another muscle that may contribute to the increase in stride length is the VM, especially in the resistance and combination conditions. These stimuli increased the VM EMG activities during the late swing phase, which may help increase knee extension and move the foot further forward. However, this point needs to be verified in future study with investigation on the knee kinematics. In addition, the LTA activity increased, although this was not significant, during both the adaptation and post-adaptation periods in subjects with incomplete SCI. This suggests that swing resistance and augmented visual feedback modulate the muscle activities from not only the ipsilateral leg but also the contralateral leg in humans with SCI. Our EMG analysis was targeted at the late stance phase to the early swing phase of the right leg, which corresponds to the early to mid stance phase of the left leg. The increase in muscle activities of LTA, although not significant, during the early stance phase of the left leg suggests that the LTA may facilitate to pull the left shank and in turn the center of mass forward over the stationary foot as an inverted pendulum (Winter, 1995). The results of this study may have clinical applications. Locomotor training has been used clinically to help people with incomplete SCI improve their gait pattern (Behrman & Harkema, 2000; Field-Fote & Roach, 2011; Harkema, 2001; Wirz et al., 2005). Our research group has provided initial evidence to support that long-term locomotor training with swing phase resistance may improve gait function in patients with incomplete SCI (Wu, Landry, Schmit, Hornby, & Yen, 2012). The results from this current study suggest that adding visual feedback (in conjunction with a single session of swing phase resistance training) may further improve training outcomes. Specifically, the combination of visual feedback and resistance training may further improve stride length and in turn improve walking speed, which may help patients return to community-based walking. More importantly to note, the increase in stride length could be greater if the two types of feedback are given simultaneously. These findings may provide suggestions in the development of gait training protocols in both clinical and lab-based settings. This study had some limitations. First, due to the small sample size used, we were not able to draw a conclusion about the impact of subjects’ injury and severity level on their responses to the various feedback conditions. Second, the investigation was focused on the spatial and temporal parameters of gait and muscle activity. Joint kinematic variables were not measured, and therefore we were not able to identify the contribution of these variables to locomotor adaptation. Third, the adaptation period was short (i.e., 7 min). It is unclear whether the results can be generalized to longer term training. Forth, in the resistance and combination conditions, a pre-tension force (4 to 5 N) was applied to the cable to prevent it from slacking during the baseline. This pre-tension force might induce some kinematic changes for some patients during the baseline. All subjects tested in this study were AIS level D (see Table 1), we do not know whether these findings can be generalized to other SCI patients with different impairment levels, such as AIS level C. Furthermore, it is unclear whether this effect can transfer to overground walking. Further studies are needed to address these questions. Lastly, the resistance created torque on the shank during the swing phase, and the torque is a function of both force and moment arm. While the same amount of force was applied to the same spot on each subject’s ankle across the loaded conditions (i.e., resistance and combination conditions), changes in hip flexion and knee extension angles during the swing phase can vary the length of the moment arm and in turn the torque. Specifically, the magnitude of torque around the hip joint will be greater during late stance and will be smaller during the mid-swing. However, the resistance was applied to the same spot on each subject’s ankle across two conditions, i.e., resistance and combination. Thus, we believe the changes in the magnitude of torque around the hip joint from the late stance to mid-swing will not systematically bias our results.

5. Conclusion We examined whether providing visual feedback in addition to swing phase resistance during locomotor adaptation could enhance the aftereffect in patients with incomplete spinal cord injury (SCI). Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

13

We found that providing visual feedback and resistance simultaneously can induce a greater and longer lasting aftereffect consisting of an increase in stride length compared to providing either augmented feedback alone. This may be due to the engagement of both implicit and explicit neural pathways for adaptation with multisensory feedback. Results from this study may be used in the development of gait training paradigms in clinical settings to improve locomotor function in humans with SCI. Acknowledgment This study is supported by Craig H. Neilsen Foundation, #124890 (Wu, M.). References Aagaard, P., Simonsen, E. B., Andersen, J. L., Magnusson, P., & Dyhre-Poulsen, P. (2002). Neural adaptation to resistance training: Changes in evoked V-wave and H-reflex responses. Journal of Applied Physiology (Bethesda, Md.: 1985), 92(6), 2309–2318. http://dx.doi.org/10.1152/japplphysiol.01185.2001. Amatachaya, S., Keawsutthi, M., Amatachaya, P., & Manimmanakorn, N. (2009). Effects of external cues on gait performance in independent ambulatory incomplete spinal cord injury patients. Spinal Cord, 47, 668–673. Anderson, K. D. (2004). Targeting recovery: Priorities of the spinal cord-injured population. Journal of Neurotrauma, 21, 1371–1383. Banz, R., Bolliger, M., Colombo, G., Dietz, V., & Lunenburger, L. (2008). Computerized visual feedback: An adjunct to roboticassisted gait training. Physical Therapy, 88(10), 1135–1145. Barbeau, H., Fung, J., Leroux, A., & Ladouceur, M. (2002). A review of the adaptability and recovery of locomotion after spinal cord injury. Progress in Brain Research, 137, 9–25. Bastian, A. J. (2008). Understanding sensorimotor adaptation and learning for rehabilitation. Current Opinion in Neurology, 21(6), 628–633. Behrman, A. L., & Harkema, S. J. (2000). Locomotor training after human spinal cord injury: A series of case studies. Physical Therapy, 80, 688–700. Beloozerova, I. N., & Sirota, M. G. (2003). Integration of motor and visual information in the parietal area 5 during locomotion. Journal of Neurophysiology, 90, 961–971. Blanchette, A., & Bouyer, L. J. (2009). Timing-specific transfer of adapted muscle activity after walking in an elastic force field. Journal of Neurophysiology, 102, 568–577. Boyd, L. A., & Winstein, C. J. (2001). Implicit motor-sequence learning in humans following unilateral stroke: The impact of practice and explicit knowledge. Neuroscience Letters, 298, 65–69. Boyd, L. A., & Winstein, C. J. (2004). Providing explicit information disrupts implicit motor learning after basal ganglia stroke. Learning & Memory, 11, 388–396. Butler, A. J., & James, K. H. (2011). Cross-modal versus within-modal recall: Differences in behavioral and brain responses. Behavioural Brain Research, 224, 387–396. Curran, T., & Keele, S. W. (1993). Attentional and nonattentional forms of sequence learning. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19, 189–202. Dietz, V., Müller, R., & Colombo, G. (2002). Locomotor activity in spinal man: Significance of afferent input from joint and load receptors. Brain, 125, 2626–2634. Field-Fote, E. C., & Roach, K. E. (2011). Influence of a locomotor training approach on walking speed and distance in people with chronic spinal cord injury: A randomized clinical trial. Physical Therapy, 91, 48–60. Fortin, K., Blanchette, A., McFadyen, B. J., & Bouyer, L. J. (2009). Effects of walking in a force field for varying durations on aftereffects and on next day performance. Experimental Brain Research, 199, 145–155. Green, T. D., & Flowers, J. H. (1991). Implicit versus explicit learning processes in a probabilistic, continuous fine-motor catching task. Journal of Motor Behavior, 23, 293–300. Harkema, S. J. (2001). Neural plasticity after human spinal cord injury: Application of locomotor training to the rehabilitation of walking. The Neuroscientist, 7, 455–468. Houldin, A., Luttin, K., & Lam, T. (2011). Locomotor adaptations and after effects to resistance during walking in individuals with spinal cord injury. Journal of Neurophysiology, 106, 247–258. Kaelin-Lang, A., Sawaki, L., & Cohen, L. G. (2005). Role of voluntary drive in encoding an elementary motor memory. Journal of Neurophysiology, 93, 1099–1103. Lam, T., Anderschitz, M., & Dietz, V. (2006). Contribution of feedback and feedforward strategies to locomotor adaptations. Journal of Neurophysiology, 95, 766–773. Lam, T., & Pearson, K. G. (2001). Proprioceptive modulation of hip flexor activity during the swing phase of locomotion in decerebrate cats. Journal of Neurophysiology, 86, 1321–1332. Lam, T., Wirz, M., Lünenburger, L., & Dietz, V. (2008). Swing phase resistance enhances flexor muscle activity during treadmill locomotion in incomplete spinal cord injury. Neurorehabilitation and Neural Repair, 22, 438–446. Malone, L. A., & Bastian, A. J. (2010). Thinking about walking: Effects of conscious correction versus distraction on locomotor adaptation. Journal of Neurophysiology, 103, 1954–1962. Morton, S. M., & Bastian, A. J. (2006). Cerebellar contributions to locomotor adaptations during splitbelt treadmill walking. The Journal of Neuroscience, 26, 9107–9116. Noreau, L., Fougeyrollas, P., Post, M., & Asano, M. (2005). Participation after spinal cord injury: The evolution of conceptualization and measurement. Journal of Neurologic Physical Therapy, 29, 147–156.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

14

S.-C. Yen et al. / Human Movement Science xxx (2014) xxx–xxx

Patton, J. L., & Mussa-Ivaldi, F. A. (2004). Robot-assisted adaptive training: Custom force fields for teaching movement patterns. IEEE Transactions on Bio-Medical Engineering, 51, 636–646. Perry, J. (1992). Gait analysis: Normal and pathological function. Thorofare, NJ: Slack. Post, M., & Noreau, L. (2005). Quality of life after spinal cord injury. Journal of Neurologic Physical Therapy, 29, 139–146. Ramnani, N. (2012). Frontal lobe and posterior parietal contributions to the cortico-cerebellar system. The Cerebellum, 11, 366–383. Reber, A. S. (1976). Implicit learning of synthetic languages: The role of instructional set. Journal of Experimental Psychology: Human Learning and Memory, 2, 88–94. Reber, P. J., & Squire, L. R. (1998). Encapsulation of implicit and explicit memory in sequence learning. Journal of Cognitive Neuroscience, 10, 248–263. Robinson, R. O., Herzog, W., & Nigg, B. M. (1987). Use of force platform variables to quantify the effects of chiropractic manipulation on gait symmetry. Journal of Manipulative and Physiological Therapeutics, 10, 172–176. Sale, D. G. (1988). Neural adaptation to resistance training. Medicine and Science in Sports and Exercise, 20(5 Suppl.), S135–S145. Shea, C. H., Wulf, G., Whitacre, C. A., & Park, J. (2001). Surfing the implicit wave. The Quarterly Journal of Experimental Psychology: Section A, 54, 841–862. Winter, D. A. (1995). Human balance and posture control during standing and walking. Gait and Posture, 3, 193–214. Wirz, M., Zemon, D. H., Rupp, R., Scheel, A., Colombo, G., Dietz, V., et al (2005). Effectiveness of automated locomotor training in patients with chronic incomplete spinal cord injury: A multicenter trial. Archives of Physical Medicine and Rehabilitation, 86, 672–680. Wu, M., Gordon, K., Kahn, J. H., & Schmit, B. D. (2011). Prolonged electrical stimulation over hip flexors increases locomotor output in human SCI. Clinical Neurophysiology, 122, 1421–1428. Wu, M., Hornby, G., Landry, J., Roth, H., & Schmit, B. (2011). A cable-driven locomotor training system for restoration of gait in human SCI. Gait and Posture, 33, 256–260. Wu, M., Landry, J. M., Schmit, B. D., Hornby, T. G., & Yen, S. C. (2012). Robotic resistance treadmill training improves locomotor function in human spinal cord injury: A pilot study. Archives of Physical Medicine and Rehabilitation, 93, 782–789. Yen, S. C., Schmit, B. D., Landry, J. M., Roth, H., & Wu, M. (2012). Locomotor adaptation to resistance during treadmill training transfers to overground walking in human SCI. Experimental Brain Research, 216, 473–482. Zeni, J. A., Richards, J. G., & Higginson, J. S. (2008). Two simple methods for determining gait events during treadmill and overground walking using kinematic data. Gait and Posture, 27, 710–714.

Please cite this article in press as: Yen, S.-C., et al. Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury. Human Movement Science (2014), http://dx.doi.org/10.1016/ j.humov.2014.03.006

Augmented multisensory feedback enhances locomotor adaptation in humans with incomplete spinal cord injury.

Different forms of augmented feedback may engage different motor learning pathways, but it is unclear how these pathways interact with each other, esp...
2MB Sizes 0 Downloads 4 Views