Effects of a New Radio Frequency Controlled Neuroprosthesis on Gait Symmetry and Rhythmicity in Patients with Chronic Hemiparesis

Transcription

1 Authors: Jeffrey M. Hausdorff, PhD Haim Ring, MD, MSc Affiliations: From the Department of Physical Therapy, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel (JMH); Movement Disorders Unit, Neurology Department, Tel-Aviv Sourasky Medical Center, Tel-Aviv, Israel (JMH); Division on Aging, Harvard Medical School, Boston, Massachusetts (JMH); Loewenstein Rehabilitation Center, Ranana, Israel (HR); and Department of Rehabilitation Medicine, Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv, Israel (HR). Correspondence: Correspondence: Haim Ring, MD, Neurological Rehabilitation Department, Loewenstein Rehabilitation Center, PO Box 3, Raanana 43100, Israel, fax: ; address: Disclosures: Portions of this work were presented at the 15th European Congress of Physical and Rehabilitation Medicine (ESPRM), May 2006, Madrid, Spain, and at the American Physical Therapy Association s Combined Sections Meeting, February 2007, Boston, Massachusetts. This work was supported in part by Ness Ltd /08/ /0 American Journal of Physical Medicine & Rehabilitation Copyright 2007 by Lippincott Williams & Wilkins DOI: /PHM.0b013e31815e6680 ORIGINAL RESEARCH ARTICLE Effects of a New Radio Frequency Controlled Neuroprosthesis on Gait Symmetry and Rhythmicity in Patients with Chronic Hemiparesis ABSTRACT Stroke Hausdorff JM, Ring H: Effects of a new radio frequency controlled neuroprosthesis on gait symmetry and rhythmicity in patients with chronic hemiparesis. Am J Phys Med Rehabil 2008;87:4 13. Objective: To assess the effects of a new neuroprosthesis (NESS L300) designed to ameliorate foot drop on gait symmetry and rhythmicity during walking. Design: Twenty-four patients (mean age: yrs) with chronic hemiparesis ( yrs) whose walking was impaired by foot drop. Subjects walked for 6 mins while wearing force-sensitive insoles, once with and once without the neuroprosthesis, in randomized order. Additional assessments with the neuroprosthesis were conducted after using the device for 4 and 8 wks. Walking speed, swing, and stride time were determined, along with a gait asymmetry index and stride time variability both markers of gait stability and fall risk. Results: While wearing the neuroprosthesis, the gait asymmetry index instantly improved by 28% (from to ) and by 45% (to ; P 0.001) after 8 wks. Stride time variability decreased by 23% immediately (from % to %) and by 33% (to %; P 0.002) after 8 wks. Walking speed improved initially by 17% (from to m/sec) and after 8 wks by 34% (to m/sec; P 0.001). Conclusions: The studied neuroprosthesis enhances gait and improves dynamic stability in chronic hemiparetic patients, supporting the idea that this is a viable treatment option in the rehabilitation of patients with foot drop. Key Words: Gait, Neuroprosthesis, Foot Drop, Stroke, Gait Variability, Asymmetry 4 Am. J. Phys. Med. Rehabil. Vol. 87, No. 1

2 Foot drop is one of the common gait disturbances caused by a stroke. An estimated 20% of all stroke survivors suffer from a foot drop. 1 It is also a common impairment in other neurological conditions such as multiple sclerosis and traumatic brain injury. Foot drop is caused by a loss of motor control over the ankle. While walking, patients with a foot drop tend to drag the foot through the swing phase and land with a flat or inverted foot during the initial contact of stance. This inefficient and unstable gait pattern results in slow walking, high energy expenditure, and increased likelihood of stumbling and falling; it may also cause pain in the joints and muscles around the hip and knee. 2 Foot drop is traditionally treated by an ankle foot orthosis (AFO). 3 Despite the limitations 3,4 of this treatment modality and the controversial evidence regarding its ability to improve hemiparetic gait, it is still the most prevalent treatment for foot drop today. 5 Functional electrical stimulation (FES) was initially introduced by Liberson and colleagues 6 as an alternative treatment for foot drop. Since then, a number of studies have demonstrated the benefits of FES for the correction of foot drop. 4,7 12 Studies that have tested the therapeutic effects of these systems have shown that FES may be effective in the rehabilitation of patients who had suffered a stroke FES systems, like those that are used to substitute for loss of function attributable to neurological damage, are often called neuroprostheses. 16 When used to correct foot drop, the stimulation of the electrodes of such systems is timed to the gait cycle, causing dorsiflexion throughout the swing phase and, thereby, reducing foot drop. Previous investigations of the efficacy of these systems generally have considered gait velocity and energy cost as the outcome measures. These studies have concluded that FES may be a useful orthotic device for a selected subset of hemiparetic patients who have maintained walking ability and whose walking is significantly impaired by foot drop. 4,7 12 However, gait speed and energy consumption may not be adequate to fully assess the impact of FES interventions on gait and functional abilities, 17 because these measures only indirectly consider dynamic balance, rhythmicity, and symmetry. None of the previous FES studies 4,7 12 have focused on these aspects of gait. Dynamic stability refers to the ability to move within a given posture without loss of balance. Gait rhythmicity and gait asymmetry are features of gait that have been associated with dynamic stability, an increased risk for falls, and diminished functional abilities Gait symmetry is an interlimb coordination parameter that is used to describe the relationship between the limbs. 24 Hemiparetic gaits exhibit asymmetries in many factors of walking. On the affected lower extremity, less time is spent in stance and single-limb stance, and more time is spent in swing. Consequently, weight bearing is unevenly distributed, and this may lead to an increased risk of falls. 2,25 Gait rhythmicity reflects the consistency of walking and the ability to maintain a steady walking pattern. Gait rhythmicity can be evaluated by measuring the stride-to-stride variability of gait timing. A more variable gait may predispose to unsteadiness and falls. 21,22,26 Evaluation of these aspects of gait may provide a measure of the degree to which a neuroprosthesis restores a physiologic gait pattern, enhances balance, and reduces the risk of falls. 27 The primary objective of this study, therefore, was to investigate the effects of the NESS L300, an FES neuroprosthesis designed to ameliorate foot drop, on walking in patients with foot drop. In particular, we tested the hypothesis that the studied neuroprosthesis enhances walking symmetry and rhythmicity in this group of patients. METHODS Participants We studied 24 patients with chronic hemiparesis. The sample size was chosen on the basis of the results of previous studies with FES devices for drop foot correction, and on the basis of a small pilot study. Subjects were recruited from two outpatient clinics in rehabilitation centers. The criteria for subject selection were (1) diagnosis of an upper motor neuron lesion, (2) chronic phase ( 6 mos after diagnosis), (3) foot drop toe drag during walking, (4) passive ankle range of motion to neutral, (5) ability to walk at least 10 m independently or with a cane, and (6) ability to follow multiplestep directions, and a score greater than 23 on the Mini-Mental State Exam (MMSE). 28 Subjects were excluded if they had a cardiac pacemaker, skin lesion at the site of the stimulation electrodes, calf muscle spasticity of more than 4 according to the modified Ashworth scale, or major depression as defined using DSM IV criteria. The FES Neuroprosthesis The neuroprosthesis (NESS L300, supplied by NESS Ltd; see Fig. 1) delivers electrical pulses to the common peroneal nerve throughout the swing phase of gait, resulting in ankle dorsiflexion to prevent foot drop. The system consists of three main components that communicate via radio frequency signals: (1) A hybrid orthosis with integrated stimulation unit and electrodes. One electrode is located over the common peroneal nerve, posterior and distal to the fibular head, and a second electrode is located over the tibialis anterior January 2008 Radio Frequency Controlled Neuroprosthesis 5

3 FIGURE 1 The NESS L300 neuroprosthesis. muscle to achieve dorsiflexion with slight eversion. The movement may be further adjusted by modifying the position of the electrodes during the fitting process. (2) A gait sensor that detects the force under the foot, using a force-sensitive resistor. (3) A miniature control unit. The orthosis ensures contact between the user s limb and the electrodes as well as reproducibility of electrode placement. Algorithms analyze the gait sensor s data to detect gait events (e.g., heel strike and toe-off) in real time. This information is then transmitted to the rest of the system to control the stimulation. A handheld computer (PDA) is used by a clinician to set the stimulation (e.g., intensity, pulse frequency) and gait parameters of the system (e.g., extended time the percentage of the stance time that the stimulation continues after heel contact). Procedures All subjects provided written informed consent, approved by the Loewenstein Rehabilitation Hospital s institutional review board. Demographic information was collected (e.g., diagnosis, age, sex, and affected side) along with medical history. The history and frequency of falls during the 2 mos before the study period were also obtained via subject recall. Immediately after fitting the neuroprosthesis and adjustment of electrode placement and stimulation parameters, each patient underwent successive gait evaluations with and without the neuroprosthesis, in a randomized order. There was a 4-wk adaptation period during which participants were instructed to increase their daily use of the neuroprosthesis according to the following: gradually increase the use of the neuroprosthesis to 1 hr by the end of the first week, to 4 hrs by the end of the second week, and to a whole day from the fourth week on. Two additional gait assessments with the neuroprosthesis were conducted after using the device for 4 and 8 wks. Under each condition, subjects walked up and down a 50-m hallway at their self-selected, usual walking speeds for 6 mins while wearing forcesensitive insoles. 29 The subjects were instructed to walk as far as they could in 6 mins while turning around each time they reached the end of the walkway. Temporal gait parameters (stride and percent swing duration average time), velocity, and physiologic cost index (PCI) were measured. The coefficient of variation (CV) of the stride time (the gait cycle duration) was calculated, using a method that quantifies the dynamics of steady-state walking and filters (removes) outliers. As described elsewhere, outliers for example, those generated by turns were automatically excluded from each subject s time series. 22,26,29 The CV assesses the stride-to-stride variability or (dys)rhythmicity of gait. Stride time variability is highly associated with the risk of falls. 21,22,26 Each subject s CV is defined as 100 (standard deviation of stride time/ mean stride time). The gait asymmetry index was measured and calculated as a marker of interlimb coordination. It was determined as follows: 100 {(swing time paretic swing time nonparetic)/ (swing time paretic swing time nonparetic)}. When the swing asymmetry index 0, gait is perfectly symmetrical, which means that both limbs are behaving identically with respect to stance and swing times. Higher scores on the asymmetry index indicate a lack of symmetry, a measure that has been associated with poor balance and a high risk for falls. 23 These two parameters, gait (stride) rhythmicity and symmetry, are the primary outcomes of this study. Average gait speed was determined by dividing the distance covered in 6 mins by 360 secs. To imitate daily life situations, average gait speed was also determined by measuring the time to walk 10 m on a carpet and over an obstacle course, using the protocol in the Emory Functional Ambulation Profile. 30 The PCI is based on the assumption that when increasing activity, the amount of energy expended by skeletal muscles is proportional to the increase of oxygen consumption of the body, and that this is proportional to the increase in heart rate (HR) caused by the studied activity. 31 The PCI was cal- 6 Hausdorff and Ring Am. J. Phys. Med. Rehabil. Vol. 87, No. 1

4 culated using the following formula: PCI change in HR from resting to steady state walking/walking speed. The final units of the PCI are beats per meter (beats/m). Heart rate was measured using a heart rate monitor (Polar Electro, Finland). To evaluate the participants acceptance of the FES neuroprosthesis, a questionnaire was filled out by each subject at the end of the study period. During the study, the participants were instructed to immediately report any fall or any other adverse event. Statistical Analysis Descriptive statistics were used to summarize patient demographics, background data, and patient questionnaires. Comparison of the number of falls between the 2-mo period before the study and the 2-mo period during the study was assessed using the Wilcoxon signed rank test for matched pairs. Six gait parameters were defined: gait rhythmicity (stride time variability); gait (swing) asymmetry index; speed during the 6-min walk; PCI; speed during the 10-m walk on a carpet; and speed during the 10-m walk on an obstacle course. A repeated-measures model, using the general linear model, was performed separately for each of these parameters to analyze the neuroprosthesis effect with four test time points: baseline with the device, baseline without the device, 4 wks later, and an additional 4 wks later with the NESS L300. Test time factor was tested for a significant trend, using the general linear model approach. If the test time factor was significant, a more detailed analysis was performed to investigate its source post hoc; we used the Dunnett multiple-comparisons test to compare the results at baseline without the device vs. results using the device at each time point. A P value less than 0.05 (two tailed) was considered statistically significant. RESULTS Subject Characteristics The average age of the subjects was yrs. Twenty (83.3%) patients were male. Twenty-one patients (87.5%) suffered from foot drop as a result of cerebral vascular accident, and the rest of the patients suffered from foot drop as a result of traumatic brain injury. Before the initiation of the study, 17 subjects used an AFO that was prescribed to each participant by his or her physician during rehabilitation. Several subjects had AFOs with special adjustments, such as AFOs with a hinge or with dorsiflexion assist moment. Three subjects used a Dictus band (OrtoPed, Canada), and four subjects did not use any assistive device for their foot drop. Fourteen patients (58.3%) had leftside hemiparesis, and 10 (41.6%) had right-side hemiparesis. The average time since diagnosis was yrs. All subjects were able to walk with the neuroprosthesis immediately after fitting. Two subjects who had a calf muscle spasticity of 4 according to the modified Ashworth scale had to use high shoes during the first week of conditioning with the neuroprosthesis. Table 1 summarizes gait at baseline, without any assistance (and at all test points). As can be seen by the first column in this table, subjects walked slowly, with marked stride-to-stride variability and asymmetry at baseline. Gait Asymmetry (Swing Asymmetry Index) Figure 2 summarizes the effects of the neuroprosthesis on gait asymmetry. The swing asymmetry index improved by 28% immediately after application of the neuroprosthesis, and it continued to improve over time, reaching a 45% change after 8 wks. The test time effect was significant (P 0.001). The measurements with the neuroprosthe- TABLE 1 Effects of the neuroprosthesis on gait Without the Neuroprosthesis (Baseline) Immediately after Application After 4 wks After 8 wks P Value (Effect of Time) Asymmetry index Stride time variability (%) min walk speed (m/sec) Walking on carpet speed (m/sec) Obstacle course speed (m/sec) Physiological cost index (beats/m) * Each of the measurements with the neuroprosthesis (immediately, after 4 wks, and after 8 wks) was significantly different from baseline values (P 0.05). January 2008 Radio Frequency Controlled Neuroprosthesis 7

5 FIGURE 2 Effects of the neuroprosthesis on gait asymmetry. Error bars reflect the standard error. w/o, without. sis were significantly different from baseline values at each time point (P 0.05), and this effect increased over time. Gait Rhythmicity (Stride Time Variability) Figure 3 shows the influence of the neuroprosthesis on gait rhythmicity as quantified by stride time variability. Initial application of the neuroprosthesis reduced stride time variability by 23%, and this measure continued to improve by 27% and 33% after 4 and 8 wks, respectively. The repeatedmeasures model yielded a significant result for the factor of test time (P 0.002). The measurements with the neuroprosthesis were significantly different from baseline (P 0.05). An example of one subject s stride time variability with and without the neuroprosthesis can be seen in Figure 4. Fall Frequency Fourteen subjects reported one or more falls during the 2 mos before the study (total 24 falls, mean ). Only two patients fell during the study period (total 2, mean ); one of them fell at night while walking without the neuroprosthesis. When using the neuroprosthesis, fall frequency decreased by 92% (P 0.001) compared with baseline values. Gait Speed There was a significant increase in gait speed in each of the three tests with the neuroprosthesis. While wearing the FES neuroprosthesis, average gait speed during the 6-min walk test initially improved by 17%, increasing to 34% after 8 wks. The test time effect was significant (P 0.001), and the Dunnett multiple-comparisons test showed that the baseline value was significantly different from each of the neuroprosthesis measurements (P 0.05). The improvement in gait speed was even more dramatic when walking over the carpet and FIGURE 3 Effects of the neuroprosthesis on stride time variability (gait rhythmicity). Error bars reflect the standard error. 8 Hausdorff and Ring Am. J. Phys. Med. Rehabil. Vol. 87, No. 1

6 FIGURE 4 Example of the effects of immediate application of the neuroprosthesis on stride time variability in one subject. the obstacle course (recall Table 1). For example, when walking over the obstacle course, speed initially increased by 24%, reaching a 44% increase after 8 wks (P 0.001). The test time effect was significant in both gait conditions (P 0.001), and the Dunnett multiple-comparisons test demonstrated significant differences between the baseline and each subsequent measure. Figure 5 illustrates the influence of the neuroprosthesis on gait speed. FIGURE 5 Effects of the neuroprosthesis on gait velocity during the 6 mins walking on the floor and the obstacle course. Bars reflect the standard error. January 2008 Radio Frequency Controlled Neuroprosthesis 9

7 Effort of Walking (PCI) As depicted in Table 1, while using the neuroprosthesis, PCI was reduced. The time effect was significant (P 0.001). The measurements after 4 and 8 wks with the neuroprosthesis were significantly different from baseline values (P 0.05). User Acceptance Table 2 summarizes the subjects acceptance of the neuroprosthesis. As can be seen in Table 2, the majority of the subjects (20 out of 24) did not need help in operating the NESS L300 after initial training, and they reported that it was not difficult to place the orthosis in the correct position using one hand. According to the participants reports (all 24), the neuroprosthesis was comfortable for all-day use, and it was also convenient for wearing in social situations. All patients described the system as being safe for use, and all but one of them expressed a desire to continue using the device. All subjects reported increases in their physical activity and greater confidence in walking on slopes and uneven surfaces. With respect to problems noted, some of the subjects mentioned that the indicator lights that flash during stimulation had attracted the attention and comments of strangers. The electrodes of the neuroprosthesis needed replacement every 2 wks. Four patients reported that it was TABLE 2 Questionnaire evaluating the subjects acceptance of the neuroprosthesis Question 1. How do you feel about continuing with use of the NESS L300? 2. How would you rate the NESS L300 against other aids to assist your gait? 3. How would you describe your walking ability since using the NESS L300? 4. How much help did you need in operating the NESS L300 (e.g., positioning the orthosis to achieve accurate movement, using the control unit, charging the system)? 5. How would you describe using the NESS L300 all day long? 6. How would you rate your confidence in performing tasks that require walking with the NESS L300? Answer and Frequency 23 Enthusiastic 0 Indifferent 1 Unenthusiastic 21 More useful 3 As useful 0 Less useful 23 Better 1 Same 0 Worse 20 I rarely needed assistance 20 Very convenient 4 I occasionally needed assistance 4 Convenient 0 I needed assistance almost each time 0 Inconvenient 21 More confident 3 No difference 0 Less confident Questions 7 12 are yes/no questions Yes No 7. Do you find the use of the NESS L300 safe? 8. Do you feel greater confidence in walking on inclines and/or uneven ground while using the NESS L300? 9. Do you feel comfortable wearing the NESS L300 in social situations? 10. Have you increased your physical activities since using the NESS L300? 11. Is the NESS L300 something you would use every day, all day? 12. Would you recommend a person with your condition to use the NESS L300? 10 Hausdorff and Ring Am. J. Phys. Med. Rehabil. Vol. 87, No. 1

8 difficult to perform this task with one hand. Two patients pointed out that it was hard to see the intensity indicator on the control unit when out in the sun. Skin irritations were not reported or observed. DISCUSSION This study investigated a new FES neuroprosthesis intended to treat foot drop. Our goal was to examine its effect on walking in patients with chronic hemiparesis who suffer from foot drop. The results support the hypothesis that the studied neuroprosthesis improves gait symmetry and rhythmicity in this group of patients. Stride time variability was lowered, and the interlimb coordination, as measured by the asymmetry index, significantly improved. Intriguingly, these improvements in gait were accompanied by a dramatic reduction in the occurrence of reported falls. This suggests that while walking with the neuroprosthesis, gait became safer and more efficient. It is, however, also possible that the introduction of a new device (the neuroprosthesis) may have caused the patients to walk more carefully, even though their gait speed increased. In addition, the retrospective assessment of fall frequency (in the initiation of the study) that was based on the patients report may have led to an overestimation of the history of falls at baseline, because subjects remembered and reported falls from longer ago. Whereas the effects on fall frequency are very intriguing, long-term prospective studies should further investigate the impact of the neuroprosthesis on falls. To our knowledge, the present study is the first to document improved gait symmetry and rhythmicity using an FES neuroprosthesis. The absence of such a report in previous studies 4,7 12 may simply be attributable to the fact that those investigations focused on other outcome measures. It is also possible that the new features implemented in the neuroprosthesis used in our study contributed to these unique outcomes. The capacity of the system to adapt in real time to gait speed and rhythm may enable better accommodation of the neuroprosthesis to changing walking patterns and the environment. Rather than simply detecting the gait events (e.g., heel-off and heel contact), the neuroprosthesis algorithms calculate moving average swing/ stance time and load, to continuously adapt to various parameters. For instance, in some patients the stimulation is extended beyond heel contact to increase ankle stability or to prevent knee snapping. The duration of this extended period is dynamically adjusted to gait speed and to changes in the temporal parameters. An additional factor that may have contributed to the results was the ability to reproduce accurate foot movement by precise positioning of the orthosis (recall Table 2). Such accurate movement is essential for balanced walking. 32 A reduction in the effort of walking while using the neuroprosthesis was also observed. Spatiotemporal asymmetry leads to increased energy expenditure. 33 Consequently, improvements in gait symmetry, as were shown in this study, may lead to reductions in the effort required for walking. 20,25 Further investigation should, however, objectively evaluate the specific contributions of these characteristics of the device to the improvements observed. In a systematic review of the orthotic effects of FES in patients with foot drop, 10 the authors conclude that FES has positive effects on walking speed and PCI. A recent study by Stein et al., 4 with a stimulator controlled by a tilt sensor, has similar conclusions. Despite these reported benefits of FES, clinical use of functional electric stimulation for the correction of foot drop is not common. 4 Insufficient clinical evidence may be a partial explanation, but another important reason for its sparse use may be ergonomic and technical problems associated with FES systems. 16,34 Among the major drawbacks reported in previously described foot drop systems are difficulties with accurate electrode positioning and attachment, cumbersome interface and assembly, and unreliable detection of gait events. 7 In a survey conducted in 98 users of noninvasive foot drop stimulators, 72% had problems finding the correct electrode positions, and 58% had difficulties with wires and footswitches. 7 Only 13% reported that they did not have problems using the device. 7 Stein et al. 4 report that 2 of 26 subjects had to use a foot sensor instead of the tilt sensor. In this case, the signal with the tilt sensor did not reliably trigger the stimulation according to the gait pattern. Taylor et al. 8 report on difficulties in finding the electrode position while using the Odstock dropped-foot stimulator. Our findings are consistent with previous studies in showing improved gait speed and a reduction in the effort required for walking while using the neuroprosthesis. Participants views regarding the device used in the present study were very positive. No major difficulties in operating the system and placing the orthosis were reported by the subjects. In addition, the radio frequency connection between the system components eliminates the need for wires. Perhaps these properties may help to pave the way for wider application in the neurorehabilitation field. Nevertheless, the use of this device has several restrictions, and it is not suitable for all patients who suffer from foot drop. The neuroprosthesis is not suitable for patients who cannot control their knee during the stance phase and, as a result, have severe knee snapping. Additionally, one of the common difficulties with the use of surface electrical January 2008 Radio Frequency Controlled Neuroprosthesis 11

9 stimulation is the tendency to develop a skin irritation. Even though we have not encountered this phenomenon in our study, it is still very important to pay constant attention to skin care during neuroprosthesis use. The present study has several limitations. The neuroprosthesis was not compared with the conventional treatment of foot drop, an AFO. Further studies should directly compare these two devices. Future investigations should also examine the effect of the neuroprosthesis with a larger sample of subjects and on patients in the acute stage of rehabilitation. The promising results of the present investigation suggest that such studies are warranted. Another potential limitation of the present study is that the protocol did not include follow-up measurements without the neuroprosthesis. It is possible that learning, motivation effects, or improvement in the patients health status played a role here, independent of the neuroprosthesis. Although this possibility cannot be completely ruled out, it is not a likely explanation for the observed results. The study cohort consisted of chronic patients whose health status was stable and who were, on average, 5.8 yrs beyond the acute stage. It seems unlikely that these results could be achieved without the effect of the neuroprosthesis. CONCLUSIONS This study demonstrates that the NESS L300 neuroprosthesis enhances gait and improves gait symmetry and rhythmicity in chronic hemiparetic patients. The findings suggest that stroke and traumatic brain injury survivors who suffer from hemiparesis that causes foot drop can gain meaningful benefits by using the neuroprosthesis on initial use, and that continued use further improves mobility. This new FES neuroprosthesis may be a viable treatment option for augmenting the rehabilitation of patients with foot drop. ACKNOWLEDGMENTS The authors thank the subjects for their time and effort. REFERENCES 1. Wade D, Wood W, Heller A, Maggs J, Langton Hewer R: Walking after stroke: measurement and recovery over the first 3 months. Scand J Rehabil Med 1987;19: Wolly SM: Characteristics of gait in hemiplegia. Top Stroke Rehabil 2001;7: Geboers JF, Drost MR, Spaans F, Kuipers H, Seelen HA: Immediate and long-term effects of ankle-foot orthosis on muscle activity during walking: a randomized study of patients with unilateral foot drop. Arch Phys Med Rehabil 2002;83: Stein RB, Chong S, Everaert DG, et al: A multicenter trial of a foot drop stimulator controlled by a tilt sensor. Neurorehabil Neural Repair 2006;20: Gok H, Kucukdeveci A, Altinkaynak H, Yavuzer G, Ergon S: Effects of an ankle foot orthoses on hemiparetic gait. Clin Rehabil 2003;17: Liberson WT, Holmquest HJ, Scott D, Dow M: Functional electrotherapy, stimulation of the peroneal nerve synchronized with the swing phase of the gait of hemiplegic patients. Arch Phys Med 1961;42: Lyons GM, Sinkjaer T, Burridge JH, Wilcox DJ: A review of portable FES-based neural orthoses for the correction of drop foot. IEEE Trans Neural Syst Rehabil Eng 2002;10: Taylor PN, Burridge J, Dunkerely A, et al: Clinical use of the Odstock dropped foot stimulator: its effect on the speed and effort of walking. Arch Phys Med Rehabil 1999;80: Taylor PN, Burridge J, Dunkerely A, et al: Clinical audit of 5 years provision of the Odstock dropped foot stimulator. Artif Organs 1999;23: Kottink AI, Oostendorp LJ, Buurke JH, Nene AV, Hermens HJ, IJzerman MJ: The orthotic effect of functional electrical stimulation on the improvement of walking in stroke patients with a dropped foot: a systematic review. Artif Organs 2004;28: Shimada Y, Matsunaga T, Misawa A, Ando S, Itoi E, Konishi N: Clinical application of peroneal nerve stimulator system using percutaneous intramuscular electrodes for correction of foot drop in hemiplegic patients. Neuromodulation 2006; 9: Weber DJ, Stein RB, Chan KM, et al: BIONic WalkAide for correcting foot drop. IEEE Trans Neural Syst Rehabil Eng 2005;13: Yan T, Hui-Chan CWY, Li LS: Functional electrical stimulation improves motor recovery of the lower extremity and walking ability of subjects with first acute stroke; a randomized placebo-controlled trial. Stroke 2005;36: Daly JJ, Roenigk K, Holcomb J, et al: A randomized controlled trial of functional neuromuscular stimulation in chronic stroke subjects. Stroke 2006;37: Robbins SM, Houghton PE, Woodbury MG, Brown JL: The therapeutic effect of functional and transcutaneous electric stimulation on improving gait speed in stroke patients: a meta-analysis. Arch Phys Med Rehabil 2006;87: Peckham PH, Knutson JS: Functional electrical stimulation for neuromuscular applications. Annu Rev Biomed Eng 2005;7: Harris JE, Eng JJ, Marigold DS, Tokuno CD, Louis CL: Relationship of balance and mobility to fall incidence in people with chronic stroke. Phys Ther 2005;85: Yavuzer G, Eser F, Karakus D, Karaoglan B, Stam HJ: The effects of balance training on gait late after stroke: a randomized controlled trial. Clin Rehabil 2006;20: Donker SF, Beek PJ: Interlimb coordination in prosthetic walking: effects of asymmetry and walking velocity. Acta Psychol (Amst) 2002;110: Sackley CM: The relationship between weight-bearing asymmetry after stroke, motor function and activities of daily living. Physiother Theory Pract 1990;6: Hausdorff JM: Gait variability: methods, modeling and meaning. J Neuroengineering Rehabil 2005;20: Springer S, Giladi N, Peretz C, Yogev G, Simon ES, Hausdorff JM: Dual-tasking effects on gait variability: the role of aging, falls, and executive function. Mov Disord 2006;21: Yogev G, Plotnik M, Peretz H, Giladi N, Hausdorff JM: Gait asymmetry in patients with Parkinson s disease and elderly fallers: when does the bilateral coordination of gait require attention. Exp Brain Res 2007;177: Nichols DS: Balance training after stroke using force platform biofeedback. Phys Ther 1997;77: Haart de M, Geurts A, Huidekoper SC, Fasotti L, Limbeek J: Recovery of standing balance in post-acute stroke patients: 12 Hausdorff and Ring Am. J. Phys. Med. Rehabil. Vol. 87, No. 1

10 a rehabilitation cohort study. Arch Phys Med Rehabil 2004; 85: Hausdorff JM, Rios D, Edelberg HK: Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil 2001;82: Yang YR, Yen JG, Wang RY, Yen LL, Lieu FK: Gait outcomes after additional backward walking training in patients with stroke: a randomized controlled trial. Clin Rehabil 2005;19: Folstein MF, Folstein SE, McHugh PR: Mini-mental state. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975;12: Sheridan PL, Solomont J, Kowall N, Hausdorff JM: Influence of executive function on locomotor function: divided attention increases gait variability in Alzheimer s disease. J Am Geriatr Soc 2003;51: Wolf SL, Catlin PA, Gage K, Gurucharri K, Roberston R, Stepahen K: Establishing the reliability and validity of measurements of walking time using the Emory functional ambulation profile. Phys Ther 1999;79: Nene AV: Physiological cost index of walking in able-bodied adolescents and adults. Clin Rehabil 1993;7: Veltink PH, Slycke P, Hemssems J, Buschman R, Bultstra G, Hermens H: Three dimensional inertial sensing of foot movements for automatic tuning of a two-channel implantable drop-foot stimulator. Med Eng Phys 2003;25: Harris-Love ML, Forrester L, Macko RF, Silver KHC, Smith GV: Hemiparetic gait parameters in overground versus treadmill walking. Neurorehabil Neural Repair 2001;15: Taylor PN, Burridge JH, Dunkerley AL, et al: Patients perceptions of the Odstock Dropped Foot Stimulator (ODFS). Clin Rehabil 1999;13: January 2008 Radio Frequency Controlled Neuroprosthesis 13