ArticlePDF Available

The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation: A systematic review

Authors:
• University of Perugia. Italy

Abstract and Figures

Objective: This review examines the utility of current active, powered, wearable lower limb exoskeletons as aids to rehabilitation in paraplegic patients with gait disorders resulting from central nervous system lesions. Methods: The PRISMA guidelines were used to review literature on the use of powered and active lower limb exoskeletons for neurorehabilitative training in paraplegic subjects retrieved in a search of the electronic databases PubMed, EBSCO, Web of Science, Scopus, ProQuest, and Google Scholar. Results: We reviewed 27 studies published between 2001 and 2014, involving a total of 144 participants from the USA, Japan, Germany, Sweden, Israel, Italy, and Spain. Seventy percent of the studies were experimental tests of safety or efficacy and 29% evaluated rehabilitative effectiveness through uncontrolled (22%) or controlled (7%) clinical trials. Conclusions: Exoskeletons provide a safe and practical method of neurorehabilitation which is not physically exhausting and makes minimal demands on working memory. It is easy to learn to use an exoskeleton and they increase mobility, improve functioning and reduce the risk of secondary injury by reinstating a more normal gait pattern. A limitation of the field is the lack of experimental methods for demonstrating the relative effectiveness of the exoskeleton in comparison with other rehabilitative techniques and technologies.
Content may be subject to copyright.
Uncorrected Author Proof
NeuroRehabilitation xx (20xx) x–xx
DOI:10.3233/NRE-151265
IOS Press
1
Review Article1
The effectiveness of powered, active lower
limb exoskeletons in neurorehabilitation:
A systematic review
2
3
4
Stefano Federicia,, Fabio Melonia, Marco Bracalentiaand Maria Laura De Filippisb
5
aDepartment of Philosophy, Social & Human Sciences and Education, University of Perugia, Italy6
bNIHR MindTech Healthcare Technology Co-operative, Institute of Mental Health, Jubilee Campus,
Nottingham, UK
7
8
Abstract.9
OBJECTIVE: This review examines the utility of current active, powered, wearable lower limb exoskeletons as aids to rehabil-
itation in paraplegic patients with gait disorders resulting from central nervous system lesions.
10
11
METHOD: The PRISMA guidelines were used to review literature on the use of powered and active lower limb exoskeletons
for neurorehabilitative training in paraplegic subjects retrieved in a search of the electronic databases PubMed, EBSCO, Web of
Science, Scopus, ProQuest, and Google Scholar.
12
13
14
RESULTS: We reviewed 27 studies published between 2001 and 2014, involving a total of 144 participants from the USA, Japan,
Germany, Sweden, Israel, Italy, and Spain. Seventy percent of the studies were experimental tests of safety or efﬁcacy and 29%
evaluated rehabilitative effectiveness through uncontrolled (22%) or controlled (7%) clinical trials.
15
16
17
CONCLUSIONS: Exoskeletons provide a safe and practical method of neurorehabilitation which is not physically exhausting
and makes minimal demands on working memory. It is easy to learn to use an exoskeleton and they increase mobility, improve
functioning and reduce the risk of secondary injury by reinstating a more normal gait pattern. A limitation of the ﬁeld is the lack
of experimental methods for demonstrating the relative effectiveness of the exoskeleton in comparison with other rehabilitative
techniques and technologies.
18
19
20
21
22
Keywords: Powered active lower limb exoskeleton, paraplegic patients, gait disorders, central nervous system lesions, neurore-
habilitation, systematic review, PRISMA
23
24
1. Introduction25
Gait disorders, classiﬁed in ICD-10 as “abnormalities26
of gait and mobility” (code R26), involve a reduction in
27
autonomy and the ability to move independently. Gait
28
disorders can result from central nervous system (CNS)
29
lesions caused by, for example, spinal cord injury (SCI),
30
cerebrovascular accident (CVA), cerebral palsy (CP),
Address for correspondence: Stefano Federici, Piazza G. Ermini
1, Perugia, 06123, Italy. Tel.: +39 3473769497; Fax: +39
0759660141; E-mail: stefano.federici@unipg.it.
or infectious diseases. When gait disorders severely 31
distort or completely abolish normal gait pattern and 32
functions (ICF, codes b770.3; b770.4) the patient is 33
usually forced to rely on a wheelchair for mobility 34
and will often require the support of a caregiver (Play- 35
ford, 2015). Over time patients may develop secondary 36
complications such as hypertension, osteoporosis, and 37
bedsores. These comorbidities severely limit an indi- 38
vidual’s ability to carry out activities of daily living 39
as well as restricting social participation, and affecting 40
quality of life and mood (Suzuki et al., 2005). 41
1053-8135/15/$35.00 © 2015 – IOS Press and the authors. All rights reserved Uncorrected Author Proof 2S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation A new neurorehabilitation therapy which is different 42 from classical therapeutic techniques has been devel-43 oped. It is referred to as neurorobotic or neuroprosthetic 44 training and uses robotic devices such as the exoskele-45 ton (Moreno et al., 2011).46 Spungen et al. (2013) stated that robotics-assisted,47 powered exoskeletons represented a relatively new 48 technology that had been shown to be a safe and effec- 49 tive method of helping individuals with motor-complete 50 paraplegia to stand and walk. Exoskeletons enable51 paralyzed individuals to overcome environmental bar- 52 riers that cannot be tackled in a wheelchair, such as 53 stairs.54 Initially, exoskeletons were used in gait disorder 55 rehabilitation to facilitate recovery of upper limb move-56 ment (Chaigneau, Arsicault, Gazeau, & Zeghloul,57 2008). The ﬁrst generation of lower limb exoskele-58 tons were passive ones (Rahman et al., 2006)—i.e., 59 the exoskeleton moves the patient’s body on a pre-60 deﬁned trajectory, regardless of what the patient is61 doing (Nef & Riener, 2012). Researchers observed that 62 users’ hip, pelvis, and leg motor patterns remained 63 unchanged when they wore a robotic gait-assisting pas-64 sive exoskeleton (Kao, Lewis, & Ferris, 2010; Lee65 et al., 2014) that reduced the muscular effort required66 (Mooney, Rouse, & Herr, 2014). This ﬁnding led to the67 development of the ﬁrst active exoskeletons (Quintero,68 Farris, & Goldfarb, 2012) that enabled patients move 69 together with the robot in the desired direction (Nef &70 Riener, 2012). 71 Exoskeletons have been used to assist patients with 72 SCI by restoring their functional abilities (Spungen73 et al., 2013), to enhance the strength and muscular 74 endurance of personnel taking part in military oper- 75 ations (Herr, 2009), and to enabled disabled people76 to participate in sport: at the 2015 Rome marathon77 two paraplegic patients ran one kilometer wearing an78 exoskeleton (Fondazione Santa Lucia, 2015). Unlike 79 the human skeleton, an exoskeleton supports body80 weight externally, allowing the user to stand com-81 pletely upright, move autonomously, and to strengthen 82 and improve coordination of voluntary movements of83 the lower limbs. This technology has greater ecologi-84 cal validity than other neurorobotic techniques such as85 robotic-assisted gait training, because patients wearing86 an exoskeleton can walk and move autonomously for87 long periods of time, and on a wide range of walking88 surfaces. Exoskeletons can also be used in a wider range89 of environments; they can be used in the workplace or90 at home, as well as in a rehabilitation space, and to help91 patients perform activities of daily living. 1.1. Purpose 92 The aim of this systematic review was to examine 93 the rehabilitative capacity of current models of active, 94 powered lower limb exoskeletons when used in para- 95 plegic patients with gait disorders resulting from central 96 nervous system lesions caused by, for example, SCIs or 97 CVAs. 98 2. Method 99 This study followed the checklist in the Pre- 100 ferred Reporting Items for Systematic Reviews and 101 Meta-Analyses (PRISMA) statement (www.prisma- 102 statement.org) (Liberati et al., 2009; Moher, Liberati, 103 Tetzlaff, Altman, & The, 2009). 104 2.1. Eligibility criteria 105 We considered English-language journal articles, 106 conference papers, and conference proceedings which 107 focused on the neurorehabilitative training of para- 108 plegic subjects using powered or active lower limb 109 exoskeletons. Studies were excluded from the analy- 110 sis if they focused on unpowered, passive, or upper 111 limb exoskeletons, were based on healthy subjects, or 112 on non-rehabilitative uses of the exoskeleton. 113 2.2. Information sources 114 Relevant publications were retrieved from searches 115 of the PubMed, EBSCO, Web of Science, Scopus, 116 ProQuest, and Google Scholar electronic databases, 117 and identiﬁed by inspecting the reference lists of rele- 118 vant review articles, editorials, and handbooks (Bogue, 119 2009; Carpino, Accoto, Tagliamonte, Ghilardi, & 120 Guglielmelli, 2013; del-Ama et al., 2012; D´ ıaz, Gil, 121 &S ´ anchez, 2011; Dietz, Nef, & Rymer, 2012; Dietz & 122 Ward, 2015; Dollar & Herr, 2007, 2008; Esquenazi & 123 Packel, 2012; Ferris, Sawicki, & Domingo, 2005; Herr, 124 2009; Hong et al., 2013; Keller & Veneman, 2013; 125 Koceska & Koceski, 2013; Low, 2011; Mohammed, 126 Amirat, & Rifai, 2012; Moreno et al., 2011; Norhaﬁzan, 127 Ghazilla, Kasi, Taha, & Hamid, 2014; Pennycott, Wyss, 128 Vallery, Klamroth-Marganska, & Riener, 2012; Jos´ eL. 129 Pons, 2010; Jos´ e L Pons & Torricelli, 2014; Viteck- 130 ova, Kutilek, & Jirina, 2013; Wall, Borg, & Palmcrantz, 131 2015; Yan, Cempini, Oddo, & Vitiello, 2015). The last 132 search was run on May 20, 2015. Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 3 2.3. Search 133 We used the following search terms in all ﬁelds: (leg134 OR (lower AND (limb* OR extremity OR body))) AND135 (power* OR active) AND (aid OR assist* OR rehab*136 OR climb stair*) AND (exoskeleton* OR wearable 137 robot*) AND (stroke* OR spinal cord* OR paraplegia*138 OR paralysis* OR disab*).139 2.4. Study selection 140 A standardized protocol was used to determine eli- 141 gibility for inclusion in the review. Publications were 142 assessed independently by authors F. M. and M. B., who 143 were blind to authorship. The ﬁrst screening was based 144 on the Abstracts of publications. A further selection 145 was made after reading the full text of the articles. Dis-146 agreements were resolved by discussion between the 147 reviewers. 148 2.5. Data collection process 149 Relevant data were extracted from the included pub-150 lications using a specially developed a data extraction151 sheet, which was piloted on ten randomly selected152 included studies and reﬁned accordingly. Data were 153 extracted from the included studies by M.B. and154 checked by F. M. Disagreements were resolved by dis- 155 cussion between the authors.156 2.6. Data items157 We extracted the following information about158 included studies: (1) year of publication; (2) country159 where the study was conducted; (3) study design (exper-160 imental trial, controlled or uncontrolled clinical trial); 161 (4) brand of exoskeleton used; (5) sample inclusion cri- 162 teria; (6) sample size; (7) sample type (SCI, stroke, not 163 available); (8) measures used in the study; (9) research164 ﬁeld; (10) publication type (journal article, conference 165 proceedings, etc.).166 3. Results 167 3.1. Study selection 168 The number of hits retrieved from each electronic 169 database is given in parentheses: EBSCO (284), Web170 of Science (88), Scopus (91), ProQuest (1119), PubMed171 (16), Google Scholar (448); a total of 2,046 publi-172 cations. Inspection and assessment of the references 173 of relevant review articles, editorials, and handbooks 174 resulted in the inclusion of an additional 24 publi- 175 cations. The search was limited to English language 176 publications and retrieved 1,795 unique records, of 177 which 1,738 were discarded after screening of the 178 Abstracts because they did not focus on our review 179 topic. The full texts of the remaining 57 publications 180 were examined in more detail. Thirty studies did not 181 meet the inclusion criteria and were excluded from the 182 analysis because the samples comprised healthy par- 183 ticipants or data about participants and experimental 184 design were not reported in the full text. 185 The ﬁnal sample comprised 27 full-text publications. 186 The process by which we selected publications for 187 inclusion in this systematic review is synthesized in a 188 ﬂow diagram based on PRISMA (Fig. 1) (Liberati et al., 189 2009). 190 3.2. Study characteristics 191 All 27 studies included in the review investigated 192 the use of active, powered lower limb exoskeletons 193 in neurorehabilitation or clinical trials in patients with 194 gait disorders. Table 1 summarizes the results of the 195 included studies according to the ten data items (see 196 sub-section 2.6) extracted from each reviewed study. 197 The included studies were published between 2001 198 and 2014 at a constant annual rate and involved a total of 199 144 participants, of whom only 4% (n= 6) were healthy 200 subjects enrolled in a control group (Sylos-Labini et al., 201 2014). The remaining participants (96%) were enrolled 202 in experimental groups comprising patients with strokes 203 (n= 51, 35%) (Kawamoto et al., 2010; Nilsson et al., 204 2014; Stein, Bishop, Stein, & Wong, 2014; Watan- 205 abe, Tanaka, Inuta, Saitou, & Yanagi, 2014), SCIs 58% 206 (n= 83) (Aach et al., 2013, 2013; Belforte, Gastaldi, & 207 Sorli, 2001; Bishop, Stein, & Wong, 2012; Esquenazi, 208 Talaty, Packel, & Saulino, 2012; Farris, Quintero, & 209 Goldfarb, 2011, 2012; Farris et al., 2014; Kolakowsky- 210 Hayner, 2013; Neuhaus et al., 2011; Quintero et al., 211 2012; Sczesny-Kaiser et al., 2013; Spungen et al., 2013; 212 Strausser & Kazerooni, 2011; Strausser, Swift, Zoss, 213 & Kazerooni, 2010; Sylos-Labini et al., 2014; Talaty, 214 Esquenazi, & Brice˜ no, 2013; Tsukahara, Hasegawa, 215 & Sankai, 2009; Tsukahara, Kawanishi, Hasegawa, & 216 Sankai, 2010; Zeilig et al., 2012), or gait disorders of 217 unspeciﬁed etiology (n= 4, 3%) (Ikehara et al., 2011; 218 Mori, Okada, & Takayama, 2006; Sanz-Merodio, Ces- 219 tari, Arevalo, & Garcia, 2012). 220 Uncorrected Author Proof 4S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Fig. 1. Four-phase ﬂow diagram of the systematic review according to PRISMA. Fifty-one percent (n= 74) of all participants were221 recruited in the USA (Bishop et al., 2012; Esquenazi et222 al., 2012; Farris et al., 2011, 2012, 2014; Kolakowsky-223 Hayner, 2013; Neuhaus et al., 2011; Quintero et al., 224 2012; Spungen et al., 2013; Stein et al., 2014; Strausser 225 & Kazerooni, 2011; Strausser et al., 2010; Talaty et al., 226 2013); 19% (n= 28) in Japan (Ikehara et al., 2011; 227 Kawamoto et al., 2010; Mori et al., 2006; Tsukahara 228 et al., 2009, 2010; Watanabe et al., 2014); 11% (n= 16) 229 in Germany (Aach et al., 2013, 2014; Sczesny-Kaiser230 et al., 2013); 6% (n= 8) in Sweden (Nilsson et al., 231 2014); 4% (n= 6) in Israel (Zeilig et al., 2012); 3%232 (n= 5) in Italy (Belforte et al., 2001; Sylos-Labini et al.,233 2014); and 0.7% (n= 1) in Spain (Sanz-Merodio et al.,234 2012). 235 Fifty-six percent (n= 15) of the publications were 236 journal articles (Aach et al., 2014; Belforte et al., 2001; 237 Bishop et al., 2012; Esquenazi et al., 2012; Farris et al., 238 2011, 2014; Mori et al., 2006; Nilsson et al., 2014; Quin- 239 tero et al., 2012; Stein et al., 2014; Sylos-Labini et al., 240 2014; Tsukahara et al., 2010; Watanabe et al., 2014;241 Zeilig et al., 2012) (Kolakowsky-Hayner, 2013); 37% 242 (n= 10) were conference papers or conference proceed- 243 ings (Farris et al., 2012; Ikehara et al., 2011; Kawamoto 244 et al., 2010; Neuhaus et al., 2011; Sanz-Merodio et al., 245 2012; Spungen et al., 2013; Strausser & Kazerooni, 246 2011; Strausser et al., 2010; Talaty et al., 2013; Tsuka- 247 hara et al., 2009); and 7% (n= 2) were book chapters 248 (Aach et al., 2013; Sczesny-Kaiser et al., 2013). 249 Fifty-six percent (n= 15) of the included publications 250 related to medical research (Aach et al., 2013, 2014; 251 Bishop et al., 2012; Esquenazi et al., 2012; Kawamoto 252 et al., 2010; Kolakowsky-Hayner, 2013; Neuhaus et al., 253 2011; Nilsson et al., 2014; Sczesny-Kaiser et al., 2013; 254 Spungen et al., 2013; Stein et al., 2014; Sylos-Labini 255 et al., 2014; Talaty et al., 2013; Watanabe et al., 2014; 256 Zeilig et al., 2012), and 44% (n= 12) to engineering 257 research (Belforte et al., 2001; Farris et al., 2011, 2012, 258 2014; Ikehara et al., 2011; Mori et al., 2006; Quintero et 259 al., 2012; Sanz-Merodio et al., 2012; Strausser & Kaze- 260 rooni, 2011; Strausser et al., 2010; Tsukahara et al., 261 2009, 2010). 262 Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 5 Table 1 Summary of the included studies Studies in alphabetic order Country Study design Exoskeleton Sample inclusion criteria Sample size Sample type Measures∗∗ Research ﬁeld Publication Type Aach et al., 2013 Germany Uncontrolled Clinical Trial HAL NA 4 SCI OBSERVATION MEASURES: Walking distance; walking speed; walking time on treadmill. OUTCOME MEASURES: 10MWT; TUG; WISCI II. Medicine Book Section Aach et al., 2014 Germany Uncontrolled Clinical Trial HAL 1) Traumatic SCI with chronic incomplete (ASIA B/C/D) or complete paraplegia (ASIA A) after lesions of the conus medullaris/cauda equine with zones of partial preservation; 2) Motor functions of hip and knee extensor and ﬂexor muscle groups to be able to trigger the exoskeleton. 8 SCI CLINICAL MEASURES: Lower extremity circumferences OUTCOME MEASURES: 10MWT; TUG; 6MWT; WISCI II; LEMS; Ashworth scale. Medicine Journal Article Belforte, Gastaldi, & Sorli, 2001 Italy Experimental Trial ARGO (Advanced Reciprocating Gait Orthosis) NA 1 SCI OBSERVATION MEASURES: Rotations and applied torques for each joint; kinematic magnitudes and exchanged forces. Engineering Journal Article Bishop, Stein, & Wong, 2012 USA Uncontrolled Clinical Trial Tibion Bionic Technologies NA 1 SCI OUTCOME MEASURES: 6MWT; BBS; TUG; 10MWT. Medicine Journal Article Esquenazi, Talaty, Packel, & Saulino, 2012 USA Experimental Trial ReWalk 1) 18 to 55 years of age; 2) Motor-complete cervical (C7-8) and thoracic (T1-T12) SCI; 3) Male and nonpregnant, nonlactating woman; 4) At least 6 months after injury; 5) Regular use of a Reciprocating Gait Orthosis (RGO) or KAFOs or able to stand using a standing device; 6) Height 160 to 190 cm 7) Weight <100 kg 12 SCI OBSERVATION MEASURES: distance walked in 6 minutes; time spent to cover 10 meters. OUTCOME MEASURES: 6MWT; 10MWT. Medicine Journal Article Uncorrected Author Proof 6S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Table 1 (Continued) Studies in alphabetic order Country Study design Exoskeleton Sample inclusion criteria Sample size Sample type Measures∗∗ Research ﬁeld Publication Type Farris et al., 2014 USA Experimental Trial Vanderbilt lower limb exoskeleton, knee–ankle–foot orthoses (KAFOs). NA 1 SCI OUTCOME MEASURES: 10MWT; TUG; 6MWT. Engineering Journal Article Farris, Quintero, & Goldfarb, 2011 USA Experimental Trial Vanderbilt lower limb exoskeleton NA 1 SCI OBSERVATION MEASURES: Hip and knee joint angle trajectories during walking; average walking speed. Engineering Journal Article Farris, Quintero, & Goldfarb, 2012 USA Experimental Trial Vanderbilt lower limb exoskeleton. NA 1 SCI OBSERVATION MEASURES: knee joint torque Engineering Conference Proceedings Ikehara et al., 2011 Japan Experimental Trial Walking Assistance Device NA 2 NA OBSERVATION MEASURES: video recording of subjects’ walking. Engineering Conference Proceedings Kawamoto et al., 2010 Japan Experimental Trial HAL NA 1 Stroke OBSERVATION MEASURES: knee joint angle Medicine Conference Proceedings Mori, Okada, & Takayama, 2006 Japan Experimental Trial ABLE system NA 1 NA OBSERVATION MEASURES: time response of the angles and electric currents of each joint. Engineering Journal Article Neuhaus et al., 2011 USA Experimental Trial MINA 1) Traumatic SCI with chronic incomplete (ASIA B/C/D) or complete paraplegia (ASIA A); 2) WISCI level 9 or higher. 2 SCI OBSERVATION MEASURES: conversation with the user during walking. Medicine Conference Proceedings Nilsson et al., 2014 Sweden Uncontrolled Clinical Trial HAL 1) Time since stroke onset of <7 weeks; 2) Able to sit on a bench with/without supervision at least ﬁve minutes; 3) Unable to walk independently; 4) Sufﬁcient postural control; 5) Ability to understanding instruction; 6) Body size compatible with the HAL suit. 8 Stroke OBSERVATION MEASURES: walking speed. OUTCOME MEASURES: NIHSS; FM-LE; BBS; TUG; 10MWT; S-COVS; FAC; FES(S); BI; FIM; VAS. Medicine Journal Article Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 7 Quintero, Farris, & Goldfarb, 2012 USA Experimental Trial Vanderbilt lower limb exoskeleton NA 1 SCI OUTCOME MEASURES: TUG. Engineering Journal Article Sanz-Merodio, Cestari, Arevalo, & Garcia, 2012 Spain Experimental Trial ATLAS NA 1 NA OBSERVATION MEASURES: joint angles Engineering Conference Proceedings Sczesny-Kaiser et al., 2013 German Uncontrolled Clinical Trial HAL NA 4 SCI OUTCOME MEASURES: fMRI; EMG. Medicine Book Section Spungen, Asselin, Fineberg, Kornfeld, & Harel, 2013 USA Uncontrolled Clinical Trial ReWalk 1) 18 to 65 years of age; 2) Motor-complete paraplegia (T1 to T12); 3) Greater than 6 months elapsed since the SCI; 4) Height 160 to 190 cm; 5) Weigh <100kg. 7 SCI OBSERVATION MEASURES: walking speed. Medicine Conference Paper Stein, Bishop, Stein, & Wong, 2014 USA Controlled Clinical Trial Bionic Leg 1) Single stroke (ischemic or hemorrhagic) causing signiﬁcant leg weakness and gait alterations at least 6 months before study entry; 2) Stroke conﬁrmation through CT or MRI; 3) Independence in household ambulation with or without facilitator. 20 Stroke OUTCOME MEASURES: 10MWT; TUG; 6MWT; 5XSST; Romberg’s test; EFAP; BBS; CAFE 40. Medicine Journal Article Strausser & Kazerooni, 2011 USA Experimental Trial eLEGS NA 5 SCI OBSERVATION MEASURES: knee angle; time practicing. Engineering Conference Proceedings Strausser, Swift, Zoss, & Kazerooni, 2010 USA Experimental Trial Human Universal Load Carrier (HULC) NA 4 SCI OBSERVATION MESURES: video recording of subjects’ walking. Engineering Conference Proceedings Sylos-Labini et al., 2014 Italy Experimental Trial MindWalker Exoskeleton 1) 18 to 45 years of age; 2) Traumatic/non- traumatic SCI; 3) At least 5 months after injury; 4) Complete lesion below T7; 5) Inability to ambulate; 6) MMSE score >26. 10 SCI OBSERVATION MEASURES: joint angles and torques. OUTCOME MEASURES: EMG. Medicine Journal Article Uncorrected Author Proof 8S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Table 1 (Continued) Studies in alphabetic order Country Study design Exoskeleton Sample inclusion criteria Sample size Sample type Measures∗∗ Research ﬁeld Publication Type Talaty, Esquenazi, & Brice˜ no, 2013 USA Experimental Trial ReWalk 1) Joint integrity/absence of fractures that prevent walking; 2) Standing alone also with facilitator; 3) A complete neurological evaluation to assess motor and physiological functioning; 4) Absence of osteoporosis (BMD >–2.5) at the right limb femoral neck and the L2 to L4 spine. 12 SCI CLINICAL MEASURES: isometric muscle strength. OBSERVATION MEASURES: video recording of subjects’ walking; walking speed. OUTCOME MEASURES: EMG. Medicine Conference Proceedings Tsukahara, Hasegawa, & Sankai, 2009 Japan Experimental Trial HAL NA 1 SCI OBSERVATION MEASURES: knee joint angle. Engineering Conference Proceedings Tsukahara, Kawanishi, Hasegawa, & Sankai, 2010 Japan Experimental Trial HAL NA 1 SCI OBSERVATION MEASURES: knee joint angle. Engineering Journal Article Watanabe, Tanaka, Inuta, Saitou, & Yanagi, 2014 Japan Controlled Clinical Trial HAL 1) Hemiparesis resulting from unilateral ischemic or hemorrhagic stroke; 2) Time since stroke onset of <6 months. 22 Stroke CLINICAL MEASURES: Isometric muscle strength. OBSERVATION MEASURES: maximum walking speed. OUTCOME MEASURES: FAC; TUG; 6MWT; SPPB; FM-LE. Medicine Journal Article Zeilig et al., 2012) Israel Experimental Trial ReWalk 1) 16 to 70 years of age; 2) Weight <100kg; 3) Height from 155 to 200 cm; 4) Complete motor impairment C7–C8 or T1–T12; 5) At least 6 months since injury; 6) Regular user of a RGO or therapeutic standing frame. 6 SCI OBSERVATION MEASURES: distance walked in 6 minutes. OUTCOME MEASURES: 10MWT; TUG. Medicine Journal Article NA = Not available: In the study inclusion criteria were not reported. All the measures’ abbreviations are explained in the Table 2. ∗∗See Table 2 for the measure reference details. Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 9 Seventy percent (n= 19) of the publications were 263 reports of experimental testing of the safety or efﬁcacy264 of a particular device (Belforte et al., 2001; Esquenazi 265 et al., 2012; Farris et al., 2011, 2012, 2014; Ikehara et266 al., 2011; Kawamoto et al., 2010; Kolakowsky-Hayner,267 2013; Mori et al., 2006; Neuhaus et al., 2011; Quin-268 tero et al., 2012; Sanz-Merodio et al., 2012; Strausser 269 & Kazerooni, 2011; Strausser et al., 2010; Sylos-Labini 270 et al., 2014; Talaty et al., 2013; Tsukahara et al., 2009, 271 2010; Zeilig et al., 2012); 22% (n= 6) were uncontrolled272 clinical trials evaluating the rehabilitative effectiveness 273 of a speciﬁc device in the context of a speciﬁc condition 274 (Aach et al., 2013, 2014; Bishop et al., 2012; Nilsson275 et al., 2014; Sczesny-Kaiser et al., 2013; Spungen et 276 al., 2013); and 7% (n= 2) were controlled clinical trials277 evaluating the rehabilitative effectiveness of a speciﬁc278 device in the context of a speciﬁc condition (Stein et al.,279 2014; Watanabe et al., 2014). 280 Three types of indicators or data were used in the281 included studies:282 Clinical indicators were used to characterize 283 baseline functioning e.g. isometric muscle284 strength (hip ﬂexion and extension, knee ﬂexion 285 and extension, trunk ﬂexion and extension, 286 ankle dorsi-/plantar ﬂexion) and lower extremity287 circumferences. 288 Outcome indicators (standardized tests, neuro-289 physiological or neuroimaging techniques) were290 used to evaluate treatment effectiveness.291 Observational data on participants’ functioning292 whilst they were wearing the exoskeleton e.g. knee293 joint angle or torque, video recordings of a partici- 294 pant walking, hip and knee joint angle trajectories 295 during walking, time spent practicing, distance 296 walked in a certain time period, etc. 297 Table 2 provides details of the indicators and the298 studies in which they were used.299 In the following sub-sections we review the evi-300 dence relating to the 14 different exoskeletons (Hybrid301 Assistive Limb, HAL; ReWalk; Vanderbilt Lower Limb 302 Exoskeleton; Human Universal Load Carrier, HULC;303 MINDWALKER Exoskeleton; Advanced Reciprocat-304 ing Gait Orthosis, ARGO; Walking Assistance Device;305 eLEGS; X1 Robotic Exoskeleton, MINA; ATLAS;306 ABLE system; Tibion Bionic Technologies; Bionic307 Leg; Ekso). We consider the type of device, partici-308 pant inclusion criteria, study objectives, indicators, and309 intervention outcomes. Because of the variability in310 study design, sample type, nature of intervention, and311 outcome indicators we evaluated the included studies312 through separate qualitative syntheses for each device 313 rather than carrying out a meta-analysis. 314 Hybrid Assistive Limb (HAL). Eight publications 315 reported on the HAL exoskeleton developed by 316 Tsukuba University, Japan and the robotics company 317 Cyberdyne (Aach et al., 2013, 2014; Kawamoto et al., 318 2010; Nilsson et al., 2014; Sczesny-Kaiser et al., 2013; 319 Tsukahara et al., 2009, 2010; Watanabe et al., 2014) 320 (Table 1). 321 The studies involved 49 participants, of whom 16 322 were enrolled in Germany, 8 in Sweden, and 25 in 323 Japan. Most of these subjects (n= 31) had hemiparesis 324 after a stroke; the remainder had a SCI at the T10 level 325 resulting in complete paraplegia (n= 10) or incomplete 326 paraplegia (n= 8). The most studies speciﬁed these 327 inclusion criteria for participants: experienced stroke or 328 SCI less than seven weeks before the start of the study; 329 able to sit on a bench for at least ﬁve minutes with or 330 without supervision; unable to walk independently due 331 to paresis of lower limbs with/without somato-sensory 332 impairment and with/without spasticity; sufﬁcient pos- 333 tural control to maintain an upright, standing position 334 with aids and/or manual support; the ability to under- 335 stand training instructions and written and oral study 336 information and the capacity to give informed consent; 337 sufﬁcient motor function in the hip and knee extensor 338 and ﬂexor muscle groups to trigger the exoskeleton; 339 body size compatible with the HAL suit. 340 In all eight studies, the main goal was to evaluate 341 the device in terms of safety, feasibility, and ability to 342 increase the motor skills of patients with chronic com- 343 plete or incomplete paraplegia. Secondary goals were 344 variable: production of standing-up and sitting motion 345 support systems for completely paraplegic patients 346 (Tsukahara et al., 2009, 2010), and in another study 347 to develop and test a method of controlling volun- 348 tary motion using bioelectrical signals generated by the 349 patient (Kawamoto et al., 2010). Two studies conducted 350 in Germany in 2103 investigated the effectiveness of the 351 HAL device (Aach et al., 2013; Sczesny-Kaiser et al., 352 2013). Aach et al. (2013, 2014) evaluated initial func- 353 tional rehabilitation of gait using the supported active 354 ROM mode of HAL. Sczesny-Kaiser et al. (2013) eval- 355 uated cortical excitability and plastic changes after a 356 three-month period of treadmill training supported by 357 HAL. Nilsson et al. (2014) assessed the safety and fea- 358 sibility of the HAL exoskeleton in a European sample. 359 Watanabe et al. (2014) compared the effectiveness of 360 the single-leg version of the HAL with conventional gait 361 training in a sample of Japanese patients with sub-acute 362 stroke. 363 Uncorrected Author Proof 10 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Table 2 Summary of the measures adopted in the studies included, ordered by largest to smallest frequency of use Measure Abbreviation Frequency Measure Study where the Measures’ Type measure was administered References Timed-Up-and-Go test TUG 9 outcome measure Aach et al., 2013. Podsiadlo & Richardson, 1991. Aach et al., 2014. Bishop et al., 2012. Farris et al., 2014. Nilsson et al., 2014. Quintero et al., 2012. Stein et al., 2014. Watanabe et al., 2014. Zeilig et al., 2012. 10-Meter Walk Test 10MWT 8 outcome measure Aach et al., 2013. Peters, Middleton, Donley, Aach et al., 2014. Blanck, & Fritz, 2014. Bishop et al., 2012. Esquenazi et al., 2012. Farris et al., 2014. Nilsson et al., 2014. Stein et al., 2014. Zeilig et al., 2012. 6-Minutes Walking Test 6MWT 6 outcome measure Aach et al., 2014. Reybrouck, 2003. Bishop et al., 2012. Esquenazi et al., 2012. Farris et al., 2014. Stein et al., 2014. Watanabe et al., 2014. Knee joint angle or torque / 6 observation measure Farris et al., 2011. / Farris et al., 2012. Kawamoto et al., 2010. Strausser & Kazerooni, 2011. Tsukahara et al., 2009. Tsukahara et al., 2010. Walking speed / 6 observation measure Aach et al., 2013. / Farris et al., 2011. Nilsson et al., 2014. Spungen et al., 2013. Talaty et al., 2013. Watanabe et al., 2014. Distance walked in a / 4 observation measure Aach et al., 2013. / certain time period Esquenazi et al., 2012. Kolakowsky-Hayner, 2013. Zeilig et al., 2012. Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 11 Berg Balance Scale BBS 3 outcome measure Bishop et al., 2012. Downs, Marquez, Nilsson et al., 2014. & Chiarelli, 2013. Stein et al., 2014. Electromyography EMG 3 outcome measure Sczesny-Kaiser et al., 2013. Mohseni Bandpei Sylos-Labini et al., 2014. et al., 2014. Talaty et al., 2013. Video recording / 3 observation measure Ikehara et al., 2011. / of subject’s walking Strausser et al., 2010. Talaty et al., 2013. Ashworth Scale / 2 outcome measure Aach et al., 2014. Pandyan et al., 1999. Kolakowsky-Hayner, 2013. Fugl-Meyer assessment FM-LE 2 outcome measure Nilsson et al., 2014. Park & Choi, 2014. for Lower Extremity Watanabe et al., 2014. Functional Ambulation FAC 2 outcome measure Nilsson et al., 2014. Mehrholz, Wagner, Rutte, Category Watanabe et al., 2014. Meissner, & Pohl, 2007. Isometric muscle strength / 2 clinical measure Talaty et al., 2013. / (hip/knee/trunk ﬂexion and Watanabe et al., 2014. extension, ankle dorsi-/ plantar ﬂexion) Time practicing / 2 observation measure Strausser & Kazerooni, 2011. / Kolakowsky-Hayner, 2013. Walking Index for SCI II WISCI II 2 outcome measure Aach et al., 2013. Ditunno et al., 2013. Aach et al., 2014. Assistive devices used / 1 observation measure Kolakowsky-Hayner, 2013. / during ambulation Barthel Index BI 1 outcome measure Nilsson et al., 2014. Cuesta-Vargas & Perez-Cruzado, 2014. Braden Scale for Predicting / 1 outcome measure Kolakowsky-Hayner, 2013. Bergstrom, Braden, Pressure Sore Risk Laguzza, & Holman, 1987. California Functional Evaluation 40 CAFE 40 1 outcome measure Stein et al., 2014. Fung et al., 1997. Clinical Outcome Variable S-COVS 1 outcome measure Nilsson et al., 2014. Andersson & Franzen, 2015. Scale, Swedish Version Conversation with the / 1 observation measure Neuhaus et al., 2011. / user during walking Emory Functional EFAP 1 outcome measure Stein et al., 2014. Wolf et al., 1999. Ambulation Proﬁle Falls-Efﬁcacy Scale, FES(S) 1 outcome measure Nilsson et al., 2014. Hellstrom, Lindmark, & Swedish Version Fugl-Meyer, 2002. Uncorrected Author Proof 12 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Table 2 (Continued) Measure Abbreviation Frequency Measure Study where the Measures’ References Type measure was administered Timed-Up-and-Go test TUG 9 outcome measure Aach et al., 2013. Podsiadlo & Richardson, 1991. Five-Times-Sit- 5XSST 1 outcome measure Stein et al., 2014. Whitney et al., 2005. to-Stand Test Functional Independence Measure FIM 1 outcome measure Nilsson et al., 2014. Saji et al., 2015. Functional Magnetic fMRI 1 outcome measure Sczesny-Kaiser et al., 2013. Buxton, 2013. Resonance Imaging Kinematic magnitudes and / 1 observation measure Belforte et al., 2001. / exchanged forces Level of functional / 1 clinical measure Kolakowsky-Hayner, 2013. / mobility Losses of balance / 1 observation measure Kolakowsky-Hayner, 2013. / Lower extremity / 1 clinical measure Aach et al., 2014. / circumference Lower Extremity LEMS 1 outcome measure Aach et al., 2014. Shin, Yoo, Jung, & Goo, 2011. Motor Score National Institutes of NIHSS 1 outcome measure Nilsson et al., 2014. Yang, Zhang, & Gao, 2014. Health Stroke Scale Number of falls / 1 observation measure Kolakowsky-Hayner, 2013. / Proprioception / 1 clinical measure Kolakowsky-Hayner, 2013. / Range of motion / 1 clinical measure Kolakowsky-Hayner, 2013. / Romberg’s Test / 1 outcome measure Stein et al., 2014. Agrawal, Carey, Hoffman, Sklare, & Schubert, 2011. Rotations and applied / 1 observation measure Belforte et al., 2001. / torques for each joint Short Physical SPPB 1 outcome measure Watanabe et al., 2014. Stookey, Katzel, Steinbrenner, Performance Battery Shaughnessy, & Ivey, 2014. Spasticity / 1 clinical measure Kolakowsky-Hayner, 2013. / Step length / 1 observation measure Kolakowsky-Hayner, 2013. / Subjective Pain Scale SPS 1 outcome measure Kolakowsky-Hayner, 2013. / Time and level / 1 observation measure Kolakowsky-Hayner, 2013. / of assistance Time response of the / 1 observation measure Mori et al., 2006. / angles and electric currents of each joint Upper and lower / 1 clinical measure Kolakowsky-Hayner, 2013. / extremity motor function Visual Analogue Scale VAS 1 outcome measure Nilsson et al., 2014. Reed & Van Nostran, 2014. [EndNote codes for generating bibliography in ‘References’: (Agrawal, Carey, Hoffman, Sklare, & Schubert, 2011; Andersson & Franzen, 2015; Bergstrom, Braden, Laguzza, & Holman, 1987; Buxton, 2013; Cuesta-Vargas & Perez-Cruzado, 2014; Ditunno et al., 2013; Downs, Marquez, & Chiarelli, 2013; Fung et al., 1997; Hellstrom, Lindmark, & Fugl-Meyer, 2002; Mehrholz, Wagner, Rutte, Meissner, & Pohl, 2007; Mohseni Bandpei et al., 2014; Pandyan et al., 1999; Park & Choi, 2014; Peters, Middleton, Donley, Blanck, & Fritz, 2014; Podsiadlo & Richardson, 1991; Reed & Van Nostran, 2014; Reybrouck, 2003; Saji et al., 2015; Shin, Yoo, Jung, & Goo, 2011; Stookey, Katzel, Steinbrenner, Shaughnessy, & Ivey, 2014; Whitney et al., 2005; Wolf et al., 1999; Yang, Zhang, & Gao, 2014)]. Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 13 The clinical, observational, and outcome data used to 364 evaluate the effectiveness of HAL varied, for example365 Sczesny-Kaiser et al. (2013) used functional mag- 366 netic resonance imaging (fMRI) and electromyography367 (EMG) to evaluate cortical excitability and plastic368 changes after a three-month period of treadmill training369 supported by HAL. 370 Taken together, these studies provide convincing 371 evidence of the safety (Nilsson et al., 2014) and effec- 372 tiveness of HAL. Paraplegic patients’ functional ability373 to walk over ground improved (Aach et al., 2013, 374 2014) and the knee angle measured during leg ﬂexion 375 increased (Kawamoto et al., 2010). Diagnostic imaging376 showed enhanced paired-pulse inhibition of somatosen- 377 sory evoked potentials in both hemispheres following378 median nerve stimulation at the wrist and reduced S1379 activation of the relevant area in both hemispheres after380 tactile stimulation of the index ﬁnger (Sczesny-Kaiser 381 et al., 2013). Finally, a gait training program using382 the single-leg version of HAL facilitated independent383 walking more effectivelythan conventional gait training 384 (Watanabe et al., 2014). 385 ReWalk. We found four papers in which the ReWalk386 exoskeleton (by ReWalk Robotics, Inc.) was tested387 (Esquenazi et al., 2012; Spungen et al., 2013; Talaty388 et al., 2013; Zeilig et al., 2012) (Table 1). These clini-389 cal trials involved 37 participants with complete SCIs390 of whom 30 were injured at the C7-T12 level and 7 391 at the T1-T12 level. All the patients were paraplegic;392 31 were recruited in the USA and 6 in Israel. The 393 inclusion criteria were: motor-complete cervical (C7- 394 8) or thoracic (T1-T12) SCI according to the American395 Spinal Injury Association (ASIA) guidelines (Ameri- 396 can Spinal Injury Association & International Medical 397 Society of Paraplegia, 2000); male sex or not being398 pregnant or lactating, age between 16 and 70 years;399 injury occurrence at least six months prior to enrolment400 in the study; regular use of reciprocating gait ortho- 401 sis (RGO) or knee-ankle-foot-orthosis (KAFOs); use402 of standing devices; and body size compatible with the403 ReWalk suit. 404 The main objectives of all studies were to assess405 the safety and tolerability of the ReWalk device.406 Various secondary objectives were also investigated,407 e.g. evaluation of the effectiveness of the ReWalk in408 enabling paraplegic patients to carry out routine ambu-409 latory functions (Esquenazi et al., 2012); investigation410 of inter-individual variability in walking kinematics411 (Talaty et al., 2013); evaluation of the number of train-412 ing sessions and level of assistance needed to achieve413 speciﬁc exoskeleton-assisted mobility skills such as414 standing, walking, and stair climbing; identiﬁcation of 415 additional indoor and outdoor mobility skills critical to 416 successful use of ReWalk in the home or community 417 (Spungen et al., 2013). 418 Taken together, these eight studies demonstrated that, 419 using the ReWalk, paraplegic patients were able to 420 achieve a level of walking proﬁciency that was close 421 to that needed for limited ambulation in an urban com- 422 munity setting (Esquenazi et al., 2012; Spungen et al., 423 2013), for example, for a distance of 100 meters (Zeilig 424 et al., 2012), and that the device enabled fundamentally 425 symmetrical gait (Talaty et al., 2013). 426 Vanderbilt Lower Limb Exoskeleton was tested in 427 four studies carried out by a team of engineers from 428 Vanderbilt University, Nashville, Tennessee, chaired 429 by Quintero (Farris et al., 2011, 2012, 2014; Quintero 430 et al., 2012) (Table 1). These clinical trials involved 431 four paraplegic patients with motor- and sensory- 432 complete SCIs at the T10 level who were recruited in 433 the USA. 434 In their ﬁrst study Farris et al. (2011) assessed the 435 impact on gait of providing assistive torque at both 436 hip and knee joints using the Vanderbilt lower limb 437 exoskeleton. A year later the research team imple- 438 mented and tested the exoskeleton’s stair ascending and 439 descending capabilities (Farris et al., 2012). Finally, 440 in 2014 they compared the mobility afforded by the 441 Vanderbilt exoskeleton and KAFOs, and the level of 442 exertion required to use the devices. The Vander- 443 bilt lower limb exoskeleton system enabled paraplegic 444 patients to walk faster (Farris et al., 2012, 2014), with 445 knee and hip joint amplitudes similar to those observed 446 in walking in people without a SCI (Farris et al., 447 2011). 448 Human Universal Load Carrier (HULC) was devel- 449 oped by Kazerooni and colleagues at Ekso Bionics, 450 USA. The HULC is designed to assist able-bodied indi- 451 viduals by powering knee extension movements so it is 452 unsurprising that only one study of its effectiveness met 453 our inclusion criteria (Strausser et al., 2010) (Table 1). 454 In this study the standard HULC, which has single- 455 acting hydraulic cylinders, was replaced with a version 456 with double-acting hydraulic cylinders that provided 457 powered ﬂexion and extension at the knee, and the 458 bracing used in the standard device was augmented 459 to provide support for patients with limited leg and 460 torso muscle control. This clinical trial involved four 461 paraplegic patients with a motor-complete SCI at the 462 T5-T10 level who were enrolled in the USA. 463 The aim of the study was to establish whether an 464 exoskeleton developed for medical use would enable 465 Uncorrected Author Proof 14 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation paraplegic people to lead a more active life and thus 466 reduce the incidence of secondary complications. The467 results conﬁrmed that when adapted for medical use 468 the HULC exoskeleton is a safe and effective method469 of improving the mobility of people who are unable to470 walk unaided.471 MINDWALKER was developed in 2009 by a Euro- 472 pean Commission-funded consortium coordinated by 473 Ilzkovitz. We reviewed only one study related to the 474 project (Sylos-Labini et al., 2014) (Table 1), a clini-475 cal trial involving ten participants recruited in Italy; six 476 were healthy and four had a SCI at the T7-L1 level. 477 The aim of this study was to quantify muscle activ-478 ityduring MINDWALKER-aided walkingina sample of 479 intactand injured patients. The data collected were EMG480 activity, joint angles, and torques. In SCI patients, EMG481 activity in the upper limb muscles was augmented when482 using the MINDWALKER, although there was little leg 483 muscleactivity.Contrarytoexpectations,however, inthe484 neurologically intact subjects, leg muscle EMG activity485 was similar or greater during exoskeleton-assisted walk- 486 ing than during normal walking on a ﬂat surface. EMG 487 waveformsvariedsigniﬁcantly across the differentwalk-488 ing conditions; variability was greatest in the hamstring489 muscles. Overall, the results are consistent with a non-490 linear reorganization of locomotor output when using491 this robotic walking device.492 Advanced Reciprocating Gait Orthosis (ARGO). 493 Only one study in our review tested the ARGO, pro-494 duced by RSLsteeper, in paraplegic patients (Belforte 495 et al., 2001) (Table 1). This was a clinical trial involving 496 one participant (single subject design) with a motor-497 complete SCI at the T3 level, and was carried out in 498 Italy. The report described the design and construction 499 of the ARGO and the experimental testing of its effects500 on locomotion in paraplegic subjects. It concluded that501 the device is ideal for rehabilitation, owing to its mod-502 ular structure and ﬂexible approach to regulating gait 503 characteristics.504 Walking Assistance Device, a close-ﬁtting walking505 assistance device with self-contained control systems, 506 is produced by Honda, Japan; it was developed and507 tested by a team of engineers led by Ikehara. Only one of508 their studies was eligible for the present review (Ikehara509 et al., 2011) (Table 1). This clinical study involved two510 participants with motor paralysis who were enrolled in511 Japan and it showed that the device could reproduce the512 power of kicking motions at ankle joints. Ikehara et al.513 (2011) claimed that persons using Walking Assistance514 Device may walk freely in both indoor and outdoor515 environments.516 eLEGS. In 2010 the Berkeley Robotics and Human 517 Engineering Laboratory, University of California, 518 directed by Kazerooni, presented the eLEGS exoskele- 519 ton developed by Strausser (Strausser & Kazerooni, 520 2011) (Table 1). 521 The clinical trial included in our review took place 522 in the USA. It tested whether the eLEGS device was 523 intuitive, and easy to learn to use in a sample of ﬁve 524 participants with no leg motion as a result of SCI or 525 ataxia. The indicators used were knee angle and time 526 spent practicing. The eLEGS human machine interface 527 (HMI) was found to be easy to use; all ﬁve sub- 528 jects quickly learnt how to use it. When walking in 529 the eLEGS the ﬁve participants had a longer interval 530 between heel lift and step than an able-bodied user. This 531 interval was shorter in experienced users: in users with 532 no experience with the device the average heel lift-to- 533 step interval was 0.859 seconds, whereas in experienced 534 users it was reduced to 0.590 seconds. 535 X1 Robotic Exoskeleton (MINA) was the outcome 536 of a collaboration between the Institute for Human and 537 Machine Cognition and the NASA Johnson Space Cen- 538 ter led by Neuhaus (Neuhaus et al., 2011) (Table 1). 539 The team carried out a clinical trial in the USA which 540 involved two participants with motor-complete SCIs 541 at the T10 level. The inclusion criteria stipulated an 542 ASIA impairment score of either A (complete) or B 543 (incomplete), and a Walking Index for Spinal Cord 544 Injury (WISCI) level 9 (9: ambulates with walker, 545 with braces and no physical assistance, 10 meters). 546 The paper presents an evaluation of the rehabilitative 547 value of the MINA exoskeleton. The cognitive effort 548 required to use MINA was evaluated qualitatively in 549 terms of how easily participants could converse with 550 the researchers whilst walking in the MINA device. 551 Static standing balance stability was assessed by ask- 552 ing participants to catch and throw a ball whilst standing 553 on both legs, using one crutch for balance. The results 554 indicated that MINA facilitated walking mobility in 555 paraplegics, allowing them to achieve speeds of up to 556 0.20 m/s. The MINA device is not physically taxing to 557 use and requires minimal cognitive effort. Cognitive 558 load was evaluated indirectly by measuring the propor- 559 tion of time for which a user was able to maintain a 560 conversation and maintain eye contact whilst walking 561 in the device. 562 ATLAS. The Centre for Automation and Robotics 563 in Spain developed the prototype of the ATLAS 564 exoskeleton to help a quadriplegic child to walk. The 565 development and main features of the device were 566 described by Sanz-Merodio et al. (2012) (Table 1). 567 Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 15 The clinical trial included in our review involved 568 one participant (single subject design), an eight-year-569 old quadriplegic Spanish girl. Experiments conﬁrmed 570 that when using the device the participant was able to571 follow the gait pattern generated by the parameterized572 trajectory generator.573 ABLE. The ABLE system (Mori et al., 2006) was 574 designed and tested in the Department of Intelli- 575 gent Systems Engineering Ibaraki University, Japan 576 (Table 1).577 The clinical trial included our review involved one 578 participant (single subject design) with motor paraly- 579 sis and was carried out in Japan. Mori and colleagues580 developed “a standing style transfer system” for a per- 581 son with disabled legs. They compared this movement582 technique with their previous system (Mori, Takayama,583 Zengo, & Nakamura, 2004). The indicators used to eval-584 uate the device were response times, joint angles, and 585 power consumption of the device, corresponding with586 the torque of each joint. The participant succeeded in587 standing up with the aid of the device; a large arm force 588 was needed in the beginning, but not after the standing 589 motion had been initiated.590 Tibion Bionic Technologies. In 2013 Tibion Bionic591 Technologies was acquired by AlterG and the exoskele-592 ton tested by the team at Columbia University (Bishop593 et al., 2012) (Table 1) is now produced by AlterG. The594 clinical trial included in our review was carried out 595 in the USA and involved one participant (single sub-596 ject design) with motor-incomplete SCI at the C5-C6 597 level. The results suggested that the device is a practical 598 and useful supplement to standard physical therapy for599 incomplete SCI. 600 Bionic Leg, which is produced by the Californian 601 company AlterG, is a powered knee orthosis for patients602 with unilateral neurological or orthopedic conditions.603 It was tested in the Department of Rehabilitation604 and Regenerative Medicine, Columbia University Col- 605 lege of Physicians and Surgeons (Stein et al., 2014)606 (Table 1).607 The clinical trial included in our review involved 20 608 participants from the USA who had hemiparesis after609 a stroke. The main inclusion criterion for this study610 was having suffered a single stroke (ischemic or hem-611 orrhagic) which had caused signiﬁcant leg weakness612 and alterations in gait alterations at least 6 months613 before entry to the study. The stroke had to have been614 conﬁrmed using computed tomography or MRI. Partic-615 ipants for whom the clinical history indicated a single616 stroke whilst the imaging studies provided evidence617 of previous asymptomatic strokes were considered eli-618 gible for participation. Participants had to be capable 619 of independent ambulation in the home (with or with- 620 out the use of unilateral assistive devices or ankle or 621 foot orthoses); the study was designed to test how the 622 Bionic Leg restored mobility in the living environments 623 to stroke survivors. The results suggested that robotic 624 knee brace therapy for ambulatory stroke patients with 625 chronic hemiparesis delivered only modest additional 626 functional beneﬁts when compared with an exercise- 627 only intervention. 628 Ekso. One publication reported on the Ekso exoskele- 629 ton developed by the robotics company Ekso Bionics 630 (Kolakowsky-Hayner, 2013) (Table 1). 631 This study involved seven participants with motor- 632 complete SCI and took place in the USA. The main 633 inclusion criterion was body size compatible with the 634 Ekso suit. The feasibility and safety of the device was 635 evaluated in a sample of SCI patients who had com- 636 pleted their initial SCI rehabilitation. The effects of 637 training with the device were evaluated in terms of time 638 tolerated, distance walked, and assistance needed whilst 639 in Ekso. The results suggested that Ekso is safe for 640 use by people with a complete thoracic SCI in a con- 641 trolled environment and may enhance mobility in those 642 without volitional lower extremity function. 643 4. Discussion and conclusions 644 We have reviewed the clinical effectiveness of var- 645 ious types of active, powered, wearable lower limb 646 exoskeletons for the rehabilitation of gait disorders 647 in paraplegic patients resulting from central ner- 648 vous system lesions caused by, for example, SCIs 649 or CVAs. 650 The most commonly studied exoskeletons were the 651 HAL (Aach et al., 2013, 2014; Kawamoto et al., 2010; 652 Nilsson et al., 2014; Sczesny-Kaiser et al., 2013; Tsuka- 653 hara et al., 2009, 2010; Watanabe et al., 2014), ReWalk 654 (Esquenazi et al., 2012; Spungen et al., 2013; Talaty 655 et al., 2013; Zeilig et al., 2012), and the Vanderbilt 656 lower limb exoskeleton (Farris et al., 2011, 2012, 2014; 657 Quintero et al., 2012). 658 Studies of exoskeleton use in neurorehabilitation 659 contexts have mostly evaluated the safety of the devices, 660 the physical and cognitive effort required to use them, 661 how easy it is to learn to use them and how effectively 662 they enhance patients’ gait. 663 The studies conﬁrmed that the HAL, Tibion Bionic 664 Technologies, and Ekso devices were safe to use in con- 665 trolled environments, and with the assistance of expert 666 Uncorrected Author Proof 16 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation professionals (i.e., physiotherapists) (Bishop et al., 667 2012; Kolakowsky-Hayner, 2013; Nilsson et al., 2014).668 We found no studies evaluating the effectiveness of 669 exoskeletons outside laboratory or clinical settings, and670 hence no evidence of their effectiveness in everyday671 living environments, for example for walking on rough672 surfaces such as a sidewalk or a staircase with uneven 673 steps. 674 The physical exertion involved in using exoskeletons 675 has been little studied. Neuhaus et al. (2011) measured676 users’ pulse rate, respiration rate, skin color, and perspi- 677 ration level when using the MINA device. Given that 678 we reviewed studies of the effectiveness of exoskele-679 tons for rehabilitation purposes, it is striking that only 680 one reported data on physical exertion. The target users681 of exoskeletons are patients who have suffered physical682 trauma, so future research should also evaluate device683 characteristics such as wearability and the relationship 684 between the weight of the device and the physical exer-685 tion required to use it.686 Information about the usability and cognitive load 687 associated with the exoskeleton was reported for MINA 688 (Neuhaus et al., 2011), Tibion Bionic Technologies689 (Bishop et al., 2012), and eLEGS (Strausser & Kaze-690 rooni, 2011). Cognitive load and usability were the691 only psychological variables investigated in the selected692 studies and in most cases data were based on sim-693 ple observations of user behavior rather than objective, 694 standardized instruments. The experimental set-up used695 by Neuhaus et al. (2011) was designed to measure users’ 696 cognitive “effort”; however the analysis was limited 697 to qualitative indicators (ability to maintain eye con-698 tact and have a conversation with the evaluator whilst 699 walking in MINA). It has been argued that interna- 700 tional scientiﬁc literature on rehabilitation pays very701 little attention to personal factors (Meloni, Federici, &702 Stella, 2011; Meloni et al., 2012) and that, in partic-703 ular, it neglects the psychological variables involved 704 in matching people with assistive technology (Corradi,705 Scherer, & Lo Presti, 2012). At present there is little706 indication of psychological variables within biopsy- 707 chosocial approach to neurorehabilitation (Williams708 & Edwards, 2003). Psychological variables should be709 taken into account alongside biological and social710 variables when considering the effectiveness of tech-711 nological aids (Federici & Scherer, 2012).712 Almost all the studies we reviewed investigated713 gait enhancement, only the two studies carried out by714 Tsukahara et al. (2009, 2010) focused solely on device715 safety. The ﬁndings converged: all the exoskeletons,716 except the Bionic Leg, were reported to facilitate717 restoration of gait patterns comparable to normal walk- 718 ing on ﬂat ground. Stein et al. (2014) found that 719 robotic therapy using Bionic Leg produced only modest 720 functional beneﬁts comparable with those of a group 721 exercise intervention. Bishop et al. (2012) assessed a 722 unilateral prosthesis similar to the Bionic Leg, namely 723 the Tibion Bionic Technologies device which was 724 developed to provide assistance and training for patients 725 with unilateral neurological or orthopedic conditions. 726 They found that this robotic device was a practical, use- 727 ful adjunct to standard training; however unlike Stein 728 et al. (2014) they did not compare the performance of 729 the group using the exoskeleton with a control group. 730 In addition to the advantages noted, we consider that 731 all the exoskeletons could be considered as ecologi- 732 cal devices, namely usable in daily living environments 733 not only in a rehabilitation setting. Using an exoskele- 734 ton can improve patients’ autonomy; simply wearing 735 the device can enable them to walk independently. 736 No other neurorehabilitative or therapeutic technique 737 or technology offers such an extraordinary potential 738 gain in autonomy. However, given that it has not been 739 shown that exoskeletons are effective in home, work, 740 and daily living environments, their ecological poten- 741 tiality remains not experimentally veriﬁed. 742 Some limitations to the neurorehabilitative use of 743 exoskeletons must be noted. First, the wearability cri- 744 teria are too restrictive; they can only be used by 745 people in speciﬁc height and weight ranges (Esquenazi 746 et al., 2012; Nilsson et al., 2014; Spungen et al., 2013; 747 Sylos-Labini et al., 2014). Second, very complex and 748 specialized training is required to enable an individual 749 to use an exoskeleton autonomously at home. Third, 750 they are still extremely expensive devices and the cost 751 is not always covered by private or public healthcare 752 systems. For instance, most European health services 753 support the use of exoskeletons as part of rehabilita- 754 tion programs delivered in specialized medical centers, 755 but not private use in an individual’s home or work- 756 place. Fourth, there is a dearth of experimental evidence 757 demonstrating that exoskeletons are more effective than 758 other rehabilitative techniques and technologies. Our 759 sample included only two studies which adopted a ran- 760 domized clinical trial design comparing exoskeleton 761 use to conventional gait training (Stein et al., 2014; 762 Watanabe et al., 2014) and the results of these two 763 studies were contradictory. Fifth, none of the publica- 764 tions we reviewed analyzed users’ experiences with an 765 exoskeleton in activities of daily living. A ﬁnal limita- 766 tion is the neglect of psychological factors; for instance 767 none of the studies we reviewed reported data on 768 Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 17 psychological reactions to trauma (post-traumatic stress 769 disorder, depression, anxiety, perception of pain, body770 image perception, etc.) (McMillan, Williams, & Bryant, 771 2003) and their potential effect on acceptance and use772 of the exoskeleton.773 Conﬂict of interest774 The authors declare that there are no conﬂicts of775 interest or funding regarding publication of this paper.776 References777 Aach, M., Cruciger, O., Sczesny-Kaiser, M., Hoffken, O., Meindl,778 R. C., Tegenthoff, M., et al. (2014). Voluntary driven exoskele-779 ton as a new tool for rehabilitation in chronic spinal cord780 injury: A pilot study. Spine Journal,14(12), 2847-2853. 781 doi:10.1016/J.Spinee.2014.03.042 782 Aach, M., Meindl, R., Hayashi, T., Lange, I., Geßmann, J., Sander, 783 A., et al. (2013). Exoskeletal Neuro-Rehabilitation in Chronic784 Paraplegic Patients – Initial Results. In J. L. Pons, D. Torricelli, &785 M. Pajaro (Eds.), Converging Clinical and Engineering Research786 on Neurorehabilitation (pp. 233-236). Berlin, DE: Springer.787 Agrawal, Y., Carey, J. P., Hoffman, H. J., Sklare, D. A., & Schubert, 788 M. C. (2011). The modiﬁed Romberg Balance Test: Normative 789 data in U.S. adults. Otology and Neurotology,32(8), 1309-1311.790 doi:10.1097/MAO.0b013e31822e5bee 791 American Spinal Injury Association, & International Medical Society792 of Paraplegia (2000). International Standards for Neurological793 and Functional Classiﬁcation of Spinal Cord Injury Patients.794 Chicago, IL: ASIA.795 Andersson, P., & Franzen, E. (2015). Effects of weight-shift train-796 ing on walking ability, ambulation, and weight distribution in797 individuals with chronic stroke: A pilot study. Topics in Stroke798 Rehabilitation. doi:10.1179/1074935715Z.00000000052799 Belforte, G., Gastaldi, L., & Sorli, M. (2001). Pneumatic active800 gait orthosis. Mechatronics,11(3), 301-323. doi:10.1016/S0957-801 4158(00)00017-9802 Bergstrom, N., Braden, B. J., Laguzza, A., & Holman, V. (1987). 803 The Braden Scale for Predicting Pressure Sore Risk. Nursing804 Research,36(4), 205-210. 805 Bishop, L., Stein, J., & Wong, C. K. (2012). Robot-aided gait train-806 ing in an individual with chronic spinal cord injury: A case807 study. Journal of Neurologic Physical Therapy,36(3), 138-143.808 doi:10.1097/NPT.0b013e3182624c87809 Bogue, R. (2009). Exoskeletons and robotic prosthetics: A review810 of recent developments. Industrial Robot,36(5), 421-427.811 doi:10.1108/01439910910980141812 Buxton, R. B. (2013). The physics of functional magnetic resonance813 imaging (fMRI). Reports on Progress in Physics,76(9), 096601.814 doi:10.1088/0034-4885/76/9/096601815 Carpino, G., Accoto, D., Tagliamonte, N. L., Ghilardi, G., &816 Guglielmelli, E. (2013). Lower limb wearable robots for phys- 817 iological gait restoration: State of the art and motivations.818 Methodology & Education for Clinical Innovation,21(2), 72-80. 819 Chaigneau, D., Arsicault, M., Gazeau, J. P., & Zeghloul, S. 820 (2008). LMS robotic hand grasp and manipulation planning (an 821 isomorphic exoskeleton approach). Robotica,26(2), 177-188. 822 doi:10.1017/S0263574707003736 823 Corradi, F., Scherer, M. J., & Lo Presti, A. (2012). Measuring the 824 Assistive Technology Match. In S. Federici & M. J. Scherer 825 (Eds.), Assistive Technology Assessment Handbook (pp. 49-65). 826 London, UK: CRC Press. 827 Cuesta-Vargas, A. I., & Perez-Cruzado, D. (2014). Relationship 828 between Barthel index with physical tests in adults with 829 intellectual disabilities. Springerplus,3(543). doi:10.1186/2193- 830 1801-3-543 831 del-Ama, A. J., Koutsou, A. D., Moreno, J. C., de-los-Reyes, 832 A., Gil-Agudo, ´ A., & Pons, J. L. (2012). Review of hybrid 833 exoskeletons to restore gait following spinal cord injury. Jour- 834 nal of Rehabilitation Research and Development,49(4), 497-514. 835 doi:10.1682/JRRD.2011.03.0043 836 D´ ıaz, I., Gil, J. J., & S´ anchez, E. (2011). Lower-Limb Robotic Reha- 837 bilitation: Literature Review and Challenges. Journalof Robotics,838 2011, 1-11. doi:10.1155/2011/759764 839 Dietz, V., Nef, T., & Rymer, W. Z. (Eds.). (2012). Neurorehabilitation 840 Technology. London, UK: Springer. 841 Dietz, V., & Ward, N. (Eds.). (2015). Oxford Textbook of Neuroreha- 842 bilitation. Oxford, UK: Oxford University Press. 843 Ditunno, J. F. J., Ditunno, P. L., Scivoletto, G., Patrick, M., Dijk- 844 ers, M., Barbeau, H., et al. (2013). The Walking Index for 845 Spinal Cord Injury (WISCI/WISCI II): Nature, metric proper- 846 ties, use and misuse. Spinal Cord,51(5), 346-355. doi:10.1038/sc. 847 2013.9 848 Dollar, A. M., & Herr, H. (2007). Active orthoses for the lower-limbs: 849 Challenges and state of the art. Paper presented at the 10th IEEE 850 International Conference on Rehabilitation Robotics, Noordwijk, 851 NL. 852 Dollar, A. M., & Herr, H. (2008). Lower Extremity Exoskele- 853 tons and Active Orthoses: Challenges and State-of-the-Art. 854 IEEE Transactions on Robotics,24(1), 144-158. doi:10.1109/ 855 TRO.2008.915453 856 Downs, S., Marquez, J., & Chiarelli, P. (2013). The Berg Balance 857 Scale has high intra- and inter-rater reliability but absolute relia- 858 bility varies across the scale: A systematic review. J Physiother,859 59(2), 93-99. doi:10.1016/S1836-9553(13)70161-9 860 Esquenazi, A., & Packel, A. (2012). Robotic-assisted gait train- 861 ing and restoration. American Journal of Physical Medicine 862 and Rehabilitation,91(11 Suppl 3), 217-227. doi:10.1097/ 863 PHM.0b013e31826bce18 864 Esquenazi, A., Talaty, M., Packel, A., & Saulino, M. (2012). The 865 ReWalk powered exoskeleton to restore ambulatory function 866 to individuals with thoracic-level motor-complete spinal cord 867 injury. American Journal of Physical Medicine and Rehabilita- 868 tion,91(11), 911-921. doi:10.1097/PHM.0b013e318269d9a3 869 Farris, R. J., Quintero, H. A., & Goldfarb, M. (2011). Prelim- 870 inary evaluation of a powered lower limb orthosis to aid 871 walking in paraplegic individuals. IEEE Transactions on Neu- 872 ral Systems and Rehabilitation Engineering,19(6), 652-659. 873 doi:10.1109/TNSRE.2011.2163083 874 Farris, R. J., Quintero, H. A., & Goldfarb, M. (2012, Aug 28- 875 Sep 1). Performance evaluation of a lower limb exoskeleton 876 for stair ascent and descent with Paraplegia. Paper presented 877 at the 34th Annual International Conference of the IEEE 878 Engineering in Medicine and Biology Society: EMBC 2012, 879 San Diego, CA. 880 Uncorrected Author Proof 18 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation Farris, R. J., Quintero, H. A., Murray, S. A., Ha, K. H., Har- 881 tigan, C., & Goldfarb, M. (2014). A preliminary assessment882 of legged mobility provided by a lower limb exoskeleton883 for persons with paraplegia. IEEE Transactions on Neu-884 ral Systems and Rehabilitation Engineering,22(3), 482-490. 885 doi:10.1109/TNSRE.2013.2268320886 Federici, S., & Scherer, M. J. (Eds.). (2012). Assistive Technology 887 Assessment Handbook. Boca Raton, FL: CRC Press.888 Ferris, D., Sawicki, G., & Domingo, A. (2005). Powered Lower Limb889 Orthoses for Gait Rehabilitation. Topics in Spinal Cord Injury890 Rehabilitation,11(2), 34-49. doi:10.1310/6GL4-UM7X-519H- 891 9JYD892 Fondazione Santa Lucia (2015, Mar 19). Maratona di Roma 2015, un893 esoscheletro hi-tech per tornare a correre. Retrieved from http://894 www.hsantalucia.it/modules.php?name=News&ﬁle=article&895 sid=989896 Fung, S., Byl, N., Melnick, M., Callahan, P., Selinger, A., Ishii,897 K., et al. (1997). Functional outcomes: The development of a898 new instrument to monitor the effectiveness of physical therapy.899 European Journal of Physical Medicine &Rehabilitation,7(2),900 31-41.901 Hellstrom, K., Lindmark, B., & Fugl-Meyer, A. (2002). The 902 Falls-Efﬁcacy Scale, Swedish version: Does it reﬂect clinically903 meaningful changes after stroke? Disability and Rehabilitation,904 24(9), 471-481. doi:10.1080/09638280110105259905 Herr, H. (2009). Exoskeletons and orthoses: Classiﬁcation, design906 challenges and future directions. Journal of Neuroengineering907 and Rehabilitation,6(1), 1-9. doi:10.1186/1743-0003-6-21 908 Hong, Y. W., King, Y.-J., Yeo, W.-H., Ting, C.-H., Chuah, Y.-D., 909 Lee, J.-V., & Chok, E.-T. (2013). Lower extremity exoskeleton:910 Review and challenges surrounding the technology and its role911 in rehabilitation of lower limbs. Australian Journal of Basic and912 Applied Sciences,7(7), 520-524.913 Ikehara, T., Nagamura, K., Ushida, T., Tanaka, E., Saegusa, S.,914 Kojima, S., & Yuge, L. (2011, Jun 29-Jul 1). Development of915 closed-ﬁtting-type walking assistance device for legs and evalua-916 tion of muscle activity. Paper presented at the IEEE International917 Conference on Rehabilitation Robotics: ICORR 2011, Zurich,918 CH.919 Kao, P.-C., Lewis, C. L., & Ferris, D. P. (2010). Invariant ankle920 moment patterns when walking with and without a robotic 921 ankle exoskeleton. Journal of Biomechanics,43(2), 203-209.922 doi:10.1016/j.jbiomech.2009.09.030923 Kawamoto, H., Taal, S., Niniss, H., Hayashi, T., Kamibayashi, K., 924 Eguchi, K., & Sankai, Y. (2010, Aug 31-Sep 4). Voluntary motion925 support control of Robot Suit HAL triggered by bioelectrical 926 signal for hemiplegia. Paper presented at the 32nd Annual Inter-927 national Conference of the IEEE Engineering in Medicine and928 Biology Society: EMBC 2010, Buenos Aires, AR.929 Keller, T., & Veneman, J. (2013). Robotics for neurorehabilitation:930 Current state and future challenges. Applied Mechanics and Mate-931 rials,245, 3-8. doi:10.4028/ www.scientiﬁc.net/MM.245.3932 Koceska, N., & Koceski, S. (2013). Review: Robot Devices for Gait933 Rehabilitation. International Journal of Computer Applications,934 62(13), 1-8. doi:10.5120/10137-4279935 Kolakowsky-Hayner, S. A. (2013). Safety and Feasibility of using the936 EksoTM Bionic Exoskeleton to Aid Ambulation after Spinal Cord 937 Injury. Journal of Spine,S4(3), 1-8. doi:10.4172/2165-7939.S4-938 003 939 Lee, H. S., Song, J., Min, K., Choi, Y.-S., Kim, S.-M., Cho, S.-R., et al.940 (2014). Short-term effects of erythropoietin on neurodevelopment941 in infants with cerebral palsy: A pilot study. Brain and Develop- 942 ment,36(9), 764-769. doi:10.1016/j.braindev.2013.11.002 943 Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gøtzsche, 944 P. C., Ioannidis, J. P. A., & Moher, D. (2009). The PRISMA 945 Statement for Reporting Systematic Reviews and Meta-Analyses 946 of Studies That Evaluate Health Care Interventions: Expla- 947 nation and Elaboration. PLoS Medicine,6(7), e1000100. 948 doi:10.1371/journal.pmed.1000100 949 Low, K. H. (2011, Aug 3-5). Robot-assisted gait rehabilitation: 950 From exoskeletons to gait systems. Paper presented at the 951 Defense Science Research Conference and Expo: DSR 2011, 952 Singapore, SG. 953 McMillan, T. M., Williams, W. H., & Bryant, R. A. (2003). Post- 954 traumatic stress disorder and traumatic brain injury: A review of 955 causal mechanisms, assessment, and treatment. In H. Williams 956 & J. J. Edwards (Eds.), Biopsychosocial Approaches in Neurore- 957 habilitation: Assessment and Management of Neuropsychiatric, 958 Mood and Behavioural Disorders (pp. 149-164). Hove, UK: Psy- 959 chology Pres. 960 Mehrholz, J., Wagner, K., Rutte, K., Meissner, D., & Pohl, M. (2007). 961 Predictive validity and responsiveness of the functional ambu- 962 lation category in hemiparetic patients after stroke. Archives 963 of Physical Medicine and Rehabilitation,88(10), 1314-1319. 964 doi:10.1016/j.apmr.2007.06.764 965 Meloni, F., Federici, S., & Stella, A. (2011). The Psychologist’s 966 Role: A Neglected Presence in the Assistive Technology Assess- 967 ment Process. In G. J. Gelderblom, M. Soede, L. Adriaens, & 968 K. Miesenberger (Eds.), Everyday Technology for Independence 969 and Care: AAATE 2011 (Vol. 29, pp. 1199-1206). Amsterdam, 970 NL: IOS Press. 971 Meloni, F.,Federici, S., Stella, A., Mazzeschi, C., Cordella, B., Greco, 972 F., & Grasso, M. (2012). The Psychologist. In S. Federici & M. J. 973 Scherer (Eds.), Assistive Technology Assessment Handbook (pp. 974 149-177). Boca Raton, FL: CRC Press. 975 Mohammed, S., Amirat, Y., & Rifai, H. (2012). Lower-Limb 976 Movement Assistance through Wearable Robots: State of 977 the Art and Challenges. Advanced Robotics,26(1-2), 1-22. 978 doi:10.1163/016918611x607356 979 Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & The, P. G. 980 (2009). Preferred Reporting Items for Systematic Reviews and 981 Meta-Analyses: The PRISMA Statement. PLoS Medicine,6(7), 982 e1000097. doi:10.1371/journal.pmed.1000097 983 Mohseni Bandpei, M. A., Rahmani, N., Majdoleslam, B., Abdol- 984 lahi, I., Ali, S. S., & Ahmad, A. (2014). Reliability of 985 surface electromyography in the assessment of paraspinal 986 muscle fatigue: An updated systematic review. Journal of 987 Manipulative and Physiological Therapeutics,37(7), 510-521. 988 doi:10.1016/j.jmpt.2014.05.006 989 Mooney, L. M., Rouse, E. J., & Herr, H. M. (2014). Autonomous 990 exoskeleton reduces metabolic cost of human walking. Jour- 991 nal of Neuroengineering and Rehabilitation,11(151), 2-5. 992 doi:10.1186/1743-0003-11-151 993 Moreno, J., Ama, A., Reyes-Guzm´ an, A., Gil-Agudo, ´ A., Ceres, 994 R., & Pons, J. (2011). Neurorobotic and hybrid manage- 995 ment of lower limb motor disorders: A review. Medical and 996 Biological Engineering and Computing,49(10), 1119-1130. 997 doi:10.1007/s11517-011-0821-4 998 Mori, Y., Okada, J., & Takayama, K. (2006). Development of 999 a standing style transfer system “ABLE” for disabled lower 1000 limbs. IEEE/ASME Transactions on Mechatronics,11(4), 372- 1001 380. doi:10.1109/TMECH.2006.878558 1002 Uncorrected Author Proof S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation 19 Mori, Y., Takayama, K., Zengo, T., & Nakamura, T. (2004). Devel- 1003 opment of straight style transfer equipment for lower limbs1004 disabled “ABLE”. Journal of Robotics and Mechatronics,16(5),1005 456-463.1006 Nef, T., & Riener, R. (2012). Three-Dimensional Multi-Degree-of- 1007 Freedom Arm Therapy Robot (ARMin). In V. Dietz, T. Nef, & W.1008 Z. Rymer (Eds.), Neurorehabilitation Technology (pp. 141-157). 1009 London, UK: Springer.1010 Neuhaus, P. D., Noorden, J. H., Craig, T. J., Torres, T., Kirschbaum,1011 J., & Pratt, J. E. (2011, Jun 29–Jul 1). Design and evaluation1012 of Mina: A robotic orthosis for paraplegics. Paper presented at 1013 the IEEE International Conference on Rehabilitation Robotics:1014 ICORR 2011, Zurich, CH.1015 Nilsson, A., Vreede, K. S., Haglund, V., Kawamoto, H., Sankai, Y.,1016 & Borg, J. (2014). Gait training early after stroke with a new1017 exoskeleton – the hybrid assistive limb: A study of safety and fea-1018 sibility. Journal of Neuroengineering and Rehabilitation,11(92),1019 1-10. doi:10.1186/1743-0003-11-921020 Norhaﬁzan, A., Ghazilla, R. A. R., Kasi, V., Taha, Z., & Hamid, B.1021 (2014). A Review on lower-Limb Exoskeleton System for Sit1022 to Stand, Ascending and Descending Staircase Motion. Applied1023 Mechanics and Materials, 541-542, 1150-1155. doi:10.4028/ 1024 www.scientiﬁc.net/AMM.541-542.11501025 Pandyan, A. D., Johnson, G. R., Price, C. I., Curless, R. H., Barnes,1026 M. P., & Rodgers, H. (1999). A review of the properties and1027 limitations of the Ashworth and modiﬁed Ashworth Scales as1028 measures of spasticity. Clinical Rehabilitation,13(5), 373-383.1029 Park, E. Y., & Choi, Y. I. (2014). Psychometric Properties of 1030 the Lower Extremity Subscale of the Fugl-Myer Assess- 1031 ment for Community-dwelling Hemiplegic Stroke Patients.1032 Journal of Physical Therapy Science,26(11), 1775-1777.1033 doi:10.1589/jpts.26.17751034 Pennycott, A., Wyss, D., Vallery, H., Klamroth-Marganska, V., &1035 Riener, R. (2012). Towards more effective robotic gait training1036 for stroke rehabilitation: A review. Journal of Neuroengineering1037 and Rehabilitation,9(65), 65. doi:10.1186/1743-0003-9-651038 Peters, D. M., Middleton, A., Donley, J. W., Blanck, E. L., & Fritz, S.1039 L. (2014). Concurrent validity of walking speed values calculated1040 via the GAITRite electronic walkway and 3 meter walk test in the1041 chronic stroke population. Physiotherapy Theory and Practice,1042 30(3), 183-188. doi:10.3109/09593985.2013.845805 1043 Playford, D. (2015). The International Classiﬁcation of Functioning,1044 Disability, and Health. In V. Dietz & N. Ward (Eds.), Oxford1045 Textbook of Neurorehabilitation (pp. 3-7). Oxford, UK: Oxford 1046 University Press.1047 Podsiadlo, D., & Richardson, S. (1991). The timed “Up & Go”: A 1048 test of basic functional mobility for frail elderly persons. Journal1049 of the American Geriatrics Society,39(2), 142-148.1050 Pons, J. L. (2010). Rehabilitation Exoskeletal Robotics. IEEE1051 Engineering in Medicine and Biology Magazine,29(3), 57-63.1052 doi:10.1109/MEMB.2010.9365481053 Pons, J. L., & Torricelli, D. (Eds.). (2014). Emerging Therapies in1054 Neurorehabilitation. Berlin, DE: Springer-Verlag.1055 Quintero, H. A., Farris, R. J., & Goldfarb, M. (2012). A Method1056 for the Autonomous Control of Lower Limb Exoskeletons for1057 Persons With Paraplegia. Journal of Medical Devices,6(4), 1-6.1058 doi:10.1115/1.4007181 1059 Rahman, T., Sample, W., Jayakumar, S., King, M. M., Wee, J. Y.,1060 Seliktar, R., et al. (2006). Passive exoskeletons for assisting limb 1061 movement. Journal of Rehabilitation Research and Development,1062 43(5), 583-590. doi:10.1682/JRRD.2005.04.00701063 Reed, M. D., & Van Nostran, W. (2014). Assessing pain intensity 1064 with the visual analog scale: A plea for uniformity. Journal of 1065 Clinical Pharmacology,54(3), 241-144. doi:10.1002/jcph.250 1066 Reybrouck, T. (2003). Clinical usefulness and limitations of the 6- 1067 minute walk test in patients with cardiovascular or pulmonary 1068 disease. Chest,123(2), 325-327. doi:10.1378/chest.123.2.325 1069 Saji, N., Kimura, K., Ohsaka, G., Higashi, Y., Teramoto, Y., Usui, 1070 M., & Kita, Y. (2015). Functional independence measure scores 1071 predict level of long-term care required by patients after stroke: 1072 A multicenter retrospective cohort study. Disability and Rehabil- 1073 itation,37(4), 331-337. doi:10.3109/09638288.2014.918195 1074 Sanz-Merodio, D., Cestari, M., Arevalo,J. C., & Garcia, E. (2012, Dec 1075 11-14). A lower-limb exoskeleton for gait assistance in quadriple- 1076 gia. Paper presented at the IEEE International Conference on 1077 Robotics and Biomimetics: ROBIO 2012, Guangzhou, CN. 1078 Sczesny-Kaiser, M., H¨ offken, O., Lissek, S., Lenz, M., Schlaffke, 1079 L., Nicolas, V., et al. (2013). Neurorehabilitation in Chronic 1080 Paraplegic Patients with the HAL®Exoskeleton – Preliminary 1081 Electrophysiological and fMRI Data of a Pilot Study.In J. L. Pons, 1082 D. Torricelli, & M. Pajaro (Eds.), Converging Clinical and Engi- 1083 neering Research on Neurorehabilitation (pp. 611-615). Berlin, 1084 DE: Springer. 1085 Shin, J. C., Yoo, J. H., Jung, T. H., & Goo, H. R. (2011). Comparison 1086 of lower extremity motor score parameters for patients with motor 1087 incomplete spinal cord injury using gait parameters. Spinal Cord,1088 49(4), 529-533. doi:10.1038/sc.2010.158 1089 Spungen, A. M., Asselin, P., Fineberg, D. B., Kornfeld, S. D., & 1090 Harel, N. Y. (2013, Apr 15-17). Exoskeletal-Assisted Walking for 1091 Persons with Motor-Complete Paraplegia. Paper presented at the 1092 STO Human Factors and Medicine Panel (HFM) Symposium, 1093 Milan, IT. 1094 Stein, J., Bishop, L., Stein, D. J., & Wong, C. K. (2014). Gait training 1095 with a robotic leg brace after stroke: A randomized controlled 1096 pilot study. American Journal of Physical Medicine and Rehabil- 1097 itation,93(11), 987-994. doi:10.1097/PHM.0000000000000119 1098 Stookey, A. D., Katzel, L. I., Steinbrenner, G., Shaughnessy, 1099 M., & Ivey, F. M. (2014). The short physical performance 1100 battery as a predictor of functional capacity after stroke. Jour- 1101 nal of Stroke and Cerebrovascular Diseases,23(1), 130-135. 1102 doi:10.1016/j.jstrokecerebrovasdis.2012.11.003 1103 Strausser, K. A., & Kazerooni, H. (2011, Sep 25-30). The devel- 1104 opment and testing of a human machine interface for a mobile 1105 medical exoskeleton. Paper presented at the IEEE/RSJ Interna- 1106 tional Conference on Intelligent Robots and Systems: IROS 2011, 1107 San Francisco, CA. 1108 Strausser, K. A., Swift, T. A., Zoss, A. B., & Kazerooni, H. (2010, 1109 Sep 12-15). Prototype medical exoskeleton for paraplegic mobil- 1110 ity: First experimental results. Paper presented at the ASME 1111 2010 Dynamic Systems and Control Conference: DSCC 2010, 1112 Cambridge, MA. 1113 Suzuki, K., Kawamura, Y., Hayashi, T., Sakurai, T., Hasegawa, Y., 1114 & Sankai, Y. (2005, Oct 10-12). Intention-based walking support 1115 for paraplegia patient. Paper presented at the IEEE International 1116 Conference on Systems, Man and Cybernetics, Waikoloa, HI. 1117 Sylos-Labini, F., La Scaleia, V., d’Avella, A., Pisotta, I., Tamburella, 1118 F., Scivoletto, G., et al. (2014). EMG patterns during assisted 1119 walking in the exoskeleton. Frontiers in Human Neuroscience,1120 8(423), 1-12. doi:10.3389/fnhum.2014.00423 1121 Talaty, M., Esquenazi, A., & Brice ˜ no, J. E. (2013, Jun 24-26). Differ- 1122 entiating ability in users of the ReWalkTM powered exoskeleton: 1123 An analysis of walking kinematics. Paper presented at the IEEE 1124 Uncorrected Author Proof 20 S. Federici et al. / The effectiveness of powered, active lower limb exoskeletons in neurorehabilitation International Conference on Rehabilitation Robotics: ICORR 1125 2013 Seattle, WA.1126 Tsukahara, A., Hasegawa, Y., & Sankai, Y. (2009, Jun 23-26).1127 Standing-up motion support for paraplegic patient with Robot1128 Suit HAL. Paper presented at the IEEE International Conference 1129 on Rehabilitation Robotics: ICORR 2009, Kyoto, JP.1130 Tsukahara, A., Kawanishi, R., Hasegawa, Y., & Sankai, Y. (2010). 1131 Sit-to-Stand and Stand-to-Sit Transfer Support for Complete1132 Paraplegic Patients with Robot Suit HAL. Advanced Robotics,1133 24(11), 1615-1638. doi:10.1163/016918610X5126221134 Viteckova, S., Kutilek, P., & Jirina, M. (2013). Wearable lower limb 1135 robotics: A review. Biocybernetics and Biomedical Engineering,1136 33(2), 96-105. doi:10.1016/J.Bbe.2013.03.0051137 Wall, A., Borg, J., & Palmcrantz, S. (2015). Clinical application1138 of the Hybrid Assistive Limb (HAL) for gait training-a sys-1139 tematic review. Frontiers in Systems Neuroscience,9(48), 1-10.1140 doi:10.3389/fnsys.2015.000481141 Watanabe, H., Tanaka, N., Inuta, T., Saitou, H., & Yanagi, H. (2014).1142 Locomotion Improvement Using a Hybrid Assistive Limb in1143 Recovery Phase Stroke Patients: A Randomized Controlled Pilot1144 Study. Archives of Physical Medicine and Rehabilitation,95(11),1145 2006-2012. doi:10.1016/J.Apmr.2014.07.002 1146 Whitney, S. L., Wrisley, D. M., Marchetti, G. F., Gee, M. A., Red-1147 fern, M. S., & Furman, J. M. (2005). Clinical measurement1148 of sit-to-stand performance in people with balance disorders: Validity of data for the Five-Times-Sit-to-Stand Test. Physical 1149 Therapy,85(10), 1034-1045. 1150 Williams, H., & Edwards, J. J. (Eds.). (2003). Biopsychosocial 1151 Approachesin Neurorehabilitation: Assessment and Management 1152 of Neuropsychiatric, Mood and Behavioural Disorders. Hove, 1153 UK: Psychology Pres. 1154 Wolf, S. L., Catlin, P. A., Gage, K., Gurucharri, K., Robertson, R., 1155 & Stephen, K. (1999). Establishing the reliability and validity 1156 of measurements of walking time using the Emory Functional 1157 Ambulation Proﬁle. Physical Therapy,79(12), 1122-1133. 1158 Yan, T., Cempini, M., Oddo, C. M., & Vitiello, N. (2015). Review 1159 of assistive strategies in powered lower-limb orthoses and 1160 exoskeletons. Robotics and Autonomous Systems,64, 120-136. 1161 doi:10.1016/j.robot.2014.09.032 1162 Yang, N., Zhang, B., & Gao, C. (2014). The baseline NIHSS score 1163 in female and male patients and short-time outcome: A study in 1164 young ischemic stroke. Journal of Thrombosis and Thrombolysis,1165 37(4), 565-570. doi:10.1007/s11239-013-0986-9 1166 Zeilig, G., Weingarden, H., Zwecker, M., Dudkiewicz, I., Bloch, A., 1167 & Esquenazi, A. (2012). Safety and tolerance of the ReWalk™ 1168 exoskeleton suit for ambulation by people with complete spinal 1169 cord injury: A pilot study.Journal of Spinal Cord Medicine,35(2), 1170 96-101. doi:10.1179/2045772312Y.0000000003 1171 ... Their main objective is to restore the patient's gait based on motor learning techniques through the reproduction of gait patterns [31]. Several studies have shown the effectiveness and limitations of these devices in neurorehabilitation [22], [32], but not in traumatic injuries such as HF. ... ... Furthermore, experimental evidence comparing the use of exoskeletons and walkers for gait rehabilitation with other types of treatment, such as conventional therapy, is almost non-existent [32]. In the case of hip fracture, there are works describing case studies, but it is claimed that there is a shortage of studies with larger sample sizes and which also include differentiated control and intervention groups [29], [38]. ... ... Among the aforementioned indexes, there are clinical studies analyzing the efficacy of exoskeletons for gait rehabilitation that compute the Barthel index. However, these studies were conducted with patients with spinal cord injury and are therefore not comparable with this work [32]. We have found no studies with hip fracture patients using exoskeletons in their rehabilitation to compare these indexes. ... Article Full-text available Hip fracture is one of the most common traumatisms associated with falls in the elderly, severely affecting the patient’s mobility and independence. In recent years, the use of robotic technology has proven to be effective in gait rehabilitation, especially for neurological disorders. However, there is a lack of research validating these devices for hip fracture in elderly patients. This paper presents the design and evaluation of a novel assistive platform for hip rehabilitation, SWalker, aimed at improving the rehabilitation of this condition. Functional validation of the SWalker platform was carried out with five healthy elderly subjects and two physiotherapists. Clinical validation was conducted with 34 patients with hip fracture. The control group ($\text {n}=24$, age$= 86.38\pm 6.16$years, 75% female) followed conventional therapy, while the intervention group ($\text{n}=10$, age$= 86.80\pm 6.32$years, 90% female) was rehabilitated using SWalker. The functional validation of the device reported good acceptability (System Usability Scale >85). In the clinical validation, the control group required 68.09±27.38 rehabilitation sessions compared to 22.60±16.75 in the intervention group ($\text{p}< 0.001$). Patients in the control group needed 120.33±53.64 days to reach ambulation, while patients rehabilitated with SWalker achieved that stage in 67.11±51.07 days ($\text{p}=0.021$). FAC and Tinetti indexes presented a larger improvement in the intervention group when compared with the control group ($\text{p}=0.007$and$\text{p}=0.01\$ , respectively). The SWalker platform can be considered an effective tool to enhance autonomous gait and shorten rehabilitation therapy in elderly hip fracture patients. This result encourages further research on robotic rehabilitation platforms for hip fracture.
... Furthermore, the therapist can provide precise stimulation by controlling the hardware of the robot. Based on such repeatability and accuracy, the robot-assisted gait-training device affects the coordination between the lower extremities in the gait of hemiplegic patients, thereby affecting postural control and adaptation and improving the patient's gait [20,21]. Furthermore, the robot hardware allows control of the treadmill, harness system, and robot frame so that various techniques can be applied to the patient based on their needs [19]. ...
... Furthermore, the robot hardware allows control of the treadmill, harness system, and robot frame so that various techniques can be applied to the patient based on their needs [19]. Such a device can prevent fatigue and injuries in therapists and provide a wide range of gait-training environments for patients [13,20]. Furthermore, the robot offers effective treatment by providing precise, repetitive motions and a joint range of motion by controlling the robot's hardware [17,22]. ...
... There was a significant improvement in 10 MWT after the 6-week intervention in all robot groups (p < 0.05) but no significant difference was observed in the non-robot group (p > 0.05). Moreover, there was a significant difference in the changes in 10 MWT among the groups (F (3,20) = 21.93, p < 0.001). ...
Article
Full-text available
This study investigated the effects of robot-assisted gait training with body weight support on gait and balance in stroke patients. The study participants comprised 24 patients diagnosed with stroke. Patients were randomly assigned to four groups of six: robot A, B, C, and non-robot. The body weight support (BWS) for the harness of the robot was set to 30% of the patient’s body weight in robot group A, 50% in robot group B, and 70% in robot group C. All experimental groups received robot-assisted gait training and general physical therapy. The non-robot group underwent gait training using a p-bar, a treadmill, and general physical therapy. The intervention was performed for 30 min a day, five times a week, for 6 weeks. All participants received the intervention after the pre-test. A post-test was performed after all of the interventions were completed. Gait was measured using a 10 m Walking test (10MWT) and the timed up and go (TUG) test. Balance was assessed using the Berg Balance Scale (BBS). Robot groups A, B, and C showed significantly better 10MWT results than did the non-robot group (p < 0.5). TUG was significantly shorter in robot groups A, B, and C than in the non-robot group (p < 0.5). The BBS scores for robot group A improved significantly more than did those for robot groups B and C and the non-robot group (p < 0.5), indicating that robot-assisted gait training with body weight support effectively improved the gait of stroke patients.
... La (ré)éducation, dès le plus jeuneâge en profitant de la plasticité cérébrale, est une des clés de la réussite de l'apprentissage permettant de reproduire une marche saine sur de longues distances. L'utilisation d'une assistance robotisée de type exosquelette, a montré qu'elle permettaità l'enfant d'acquérir une marche proche de celle de la marche saine après les phases d'apprentissage (Federici et al., 2015, Lajeunesse et al., 2015, Miller et al., 2016, Wall et al., 2015 pour certaines pathologies modérées. L'objectifà terme est de développer un prototype d'exosquelette personnalisé dont l'anthropométrie serait adaptéeà l'enfantétudié. ...
... Le couple obtenu T torsion a ensuiteété ajouté au couple calculé par TSID afin de déterminer couple d'impédance utilisé pour calculer l'accélération correspondantè a l'aide d'un algorithme de corps articulé (ABA) issu de la librairie Pinocchio(Carpentier et al., 2019 ; Carpentier et al., 2015 Carpentier et al., -2021 calculant la dynamique, c'est-à-dire les accélérations des articulations en fonction de l'état actuel et de l'actionnement du modèle. Cette fonction est mise en oeuvre en se basant surFeatherstone, 2014. ...
... Exoskeletons increase mobility, improve the limb's motor function, and restore the regular walking pattern [19]. Israel)-ReWalk is a wearable robotic exoskeleton that provides powered hip and knee motion to enable individuals with spinal cord injury (SCI); REX (Rex Bionics PLC, London, UK)-mission is to develop technology in the field of autonomous mobile robotics, role in uprighting and walking training for patients with complete or incomplete C4-L5 injury; HAL (Hybrid Assistive Limb)-an exoskeleton for uprighting and training is walking; Ekso (Ekso Bionics, Richmond CA, USA)-an exoskeleton for training walking; Indego (Parker Hannifin, OH, USA)-The Indego is a powered hip-knee exoskeleton for gait training by motion. ...
... Exoskeletons increase mobility, improve the limb's motor function, and restore the regular walking pattern [19]. ...
Article
Full-text available
Spinal cord injuries (SCIs) have major consequences on the patient's health and life. Voluntary muscle paralysis caused by spinal cord damage affects the patient's independence. Following SCI, an irreversible motor and sensory deficit occurs (spasticity, muscle paralysis, atrophy, pain, gait disorders, pain). This pathology has implications on the whole organism: on the osteoarticular, muscular, cardiovascular, respiratory, gastrointestinal, genito-urinary, skin, metabolic disorders, and neuro-psychic systems. The rehabilitation process for a subject having SCIs can be considered complex, since the pathophysiological mechanism and biochemical modifications occurring at the level of spinal cord are not yet fully elucidated. This review aims at evaluating the impact of robotic-assisted rehabilitation in subjects who have suffered SCI, both in terms of regaining mobility as a major dysfunction in patients with SCI, but also in terms of improving overall fitness and cardio-vascular function, respiratory function, as well as the gastrointestinal system, bone density and finally the psychosocial issues, based on multiple clinical trials, and pilot studies. The researched literature in the topic revealed that in order to increase the chances of neuro-motor recovery and to obtain satisfactory results, the combination of robotic therapy, a complex recovery treatment and specific medication is one of the best decisions. Furthermore, the use of these exoskeletons facilitates better/greater autonomy for patients, as well as optimal social integration.
... However, the outcomes attained with ambulatory exoskeletons are still controversial. Published studies and reviews show considerable differences among protocols, targeted populations and variables analyzed, in addition to the specific differences among exoskeletons (number of joints, type of actuators and controllers, among others) [3,4]. Recent research claims that robot-assisted walking arises from the interaction between the human body, driven by the central neural system (CNS) through the muscles, neural loops, reflex mechanisms and the mechanical structure of each exoskeleton, driven by the controller through the joint actuators and sensors [5]. ...
Article
Full-text available
Nowadays, robotic technology for gait training is becoming a common tool in rehabilitation hospitals. However, its effectiveness is still controversial. Traditional control strategies do not adequately integrate human intention and interaction and little is known regarding the impact of exoskeleton control strategies on muscle coordination, physical effort, and user acceptance. In this article, we benchmarked three types of exoskeleton control strategies in a sample of seven healthy volunteers: trajectory assistance (TC), compliant assistance (AC), and compliant assistance with EMG-Onset stepping control (OC), which allows the user to decide when to take a step during the walking cycle. This exploratory study was conducted within the EUROBENCH project facility. Experimental procedures and data analysis were conducted following EUROBENCH's protocols. Specifically, exoskeleton kinematics, muscle activation, heart and breathing rates, skin conductance, as well as user-perceived effort were analyzed. Our results show that the OC controller showed robust performance in detecting stepping intention even using a corrupt EMG acquisition channel. The AC and OC controllers resulted in similar kinematic alterations compared to the TC controller. Muscle synergies remained similar to the synergies found in the literature, although some changes in muscle contribution were found, as well as an overall increase in agonist-antagonist co-contraction. The OC condition led to the decreased mean duration of activation of synergies. These differences were not reflected in the overall physiological impact of walking or subjective perception. We conclude that, although the AC and OC walking conditions allowed the users to modulate their walking pattern, the application of these two controllers did not translate into significant changes in the overall physiological cost of walking nor the perceived experience of use. Nonetheless, results suggest that both AC and OC controllers are potentially interesting approaches that can be explored as gait rehabilitation tools. Furthermore, the INTENTION project is, to our knowledge, the first study to benchmark the effects on human-exoskeleton interaction of three different exoskeleton controllers, including a new EMG-based controller designed by us and never tested in previous studies, which has made it possible to provide valuable third-party feedback on the use of the EUROBENCH facility and testbed, enriching the apprenticeship of the project consortium and contributing to the scientific community.
... Modern wearable robotic exoskeletons are powerful emerging technologies that are versatile enough for personalized patient-centered care, can be specifically designed to deliver multicomponent exercise interventions, and have been extremely beneficial in neurologically impaired clinical populations [8,9]. However, the efficacy of robotic exoskeleton-based exercise intervention in a community setting is yet to be assessed in any population, especially the older adult population. ...
Article
Full-text available
Background Despite the benefits of physical activity for healthy physical and cognitive aging, 35% of adults over the age of 75 in the United States are inactive. Robotic exoskeleton-based exercise studies have shown benefits in improving walking function, but most are conducted in clinical settings with a neurologically impaired population. Emerging technology is starting to enable easy-to-use, lightweight, wearable robots, but their impact in the otherwise healthy older adult population remains mostly unknown. For the first time, this study investigates the feasibility and efficacy of using a lightweight, modular hip exoskeleton for in-community gait training in the older adult population to improve walking function. Methods Twelve adults over the age of 65 were enrolled in a gait training intervention involving twelve 30-min sessions using the Gait Enhancing and Motivating System for Hip in their own senior living community. Results Performance-based outcome measures suggest clinically significant improvements in balance, gait speed, and endurance following the exoskeleton training, and the device was safe and well tolerated. Gait speed below 1.0 m/s is an indicator of fall risk, and two out of the four participants below this threshold increased their self-selected gait speed over 1.0 m/s after intervention. Time spent in sedentary behavior also decreased significantly. Conclusions This intervention resulted in greater improvements in speed and endurance than traditional exercise programs, in significantly less time. Together, our results demonstrated that exoskeleton-based gait training is an effective intervention and novel approach to encouraging older adults to exercise and reduce sedentary time, while improving walking function. Future work will focus on whether the device can be used independently long-term by older adults as an everyday exercise and community-use personal mobility device. Trial registration This study was retrospectively registered with ClinicalTrials.gov (ID: NCT05197127).
... √ √ [21] Compliant joint mechanisms to ensure user's safety and comfort. √ √ [22] Application of wearable lower limb exoskeletons for rehabilitation of patients with gait disorders. ...
Article
Full-text available
Wearableassistive devices such as passive exoskeletonhavebeen recognized as one of the effective solutions to assist people inindustrialwork, rehabilitation, elderly care, military and sports. Thedesign and development of a passive exoskeletonthat emphasizes on satisfying and fulfilling users’ requirements and users’ experience areessential to ensure the device remains competitive in the global market. Agood user experience of using an exoskeletonstimulates users’ satisfaction, as contemporary users are not only considering basic functional features but alsofascinated by perception values such as aesthetics and enjoyment. Themain purpose of this article is to review the critical factors that are influencing user experience before, during and after utilizing a passive exoskeleton. The authors had searched relevant articles from academic databases such as Google Scholar, Scopus and Web of Science as well as free Google search for the publicationperiod from 2001to 2021. Several search keywords were used such as ‘passive exoskeleton +user experience’, ‘passive exoskeleton + industry’, ‘passive exoskeleton + rehabilitation’, ‘passive exoskeleton + military’, ‘passive exoskeleton + sports’,‘passive exoskeleton + sit-stand’, and passive exoskeleton + walking’. This online search found that a total of 236 articles related to the application of passive exoskeleton in the area of industry, rehabilitation, military and sports. Out of this, 81 articles were identified as significant references and examined thoroughly to prepare the essence of this paper. Based on thesearticles, the authors revealed that the engineering design, usability, flexibility, safety and ergonomics, aesthetics, accessibility, purchase cost, after-sales service and sustainability are the critical factorsthatare influencing user experience whenemployingpassive exoskeleton.
... A associação com a terapia robótica talvez possa otimizar os resultados por proporcionar maior número de repetições nos movimentos realizados e complementar o treino funcional individualizado, atuando como importante método de neuroreabilitação. 9 O caso descrito é motivador para que estudos com um maior número de pacientes e com o enfoque no programa de reabilitação em modelo de internação sejam realizados. ...
Article
Full-text available
There is little information in the medical literature on the rehabilitation of patients with GuillainBarre Syndrome (GBS). There are clinical studies that demonstrate the effectiveness of a rehabilitation program using an interdisciplinary team, but without well-defined protocols and only performed on an outpatient basis. This case report aims to describe the evolution of a patient with GBS during the intensive multidisciplinary inpatient rehabilitation program, discussing the therapeutic possibilities for rehabilitation of the disease
Chapter
Over the past decade, overground robotic exoskeletons have emerged as promising technologies that can be integrated into the rehabilitation process to help individuals maintain or regain neuromuscular health following neurological injury. Early studies suggest that individuals recovering from stroke, spinal cord injury, and other neurological conditions can benefit from the use of exoskeletons, either alone or as a complement to traditional rehabilitation strategies, to improve mobility and independence. Due to the broad range of impairments observed after neurological injury, clinicians should consider different types of exoskeletons to best suit the goals of each patient. The use of these exoskeletons as clinical tools also requires clinicians to understand how to operate and monitor the device, to identify which patient population(s) are appropriate and how they may benefit from the device in rehabilitation, and the limitations and safety measures required for each device. More research in this field, including large-scale clinical trials to assess the therapeutic benefits and limitations of exoskeletons, is required to achieve a greater understanding of how to optimize the use of these devices in the clinic and for personal mobility.
Article
In this study, we focused on the design and control of a single DoF laterally supported knee exoskeleton robot designed to improve the load-carrying and strength capacity of healthy individuals, especially soldiers and workers. First, a nonlinear second-order differential model of the robotic knee orthosis was produced. Then, a single degree of freedom exoskeleton robot was designed and manufactured. An interactive motion control method based on the relative angle measurement principle between the knee joint of the user and the exo-suit knee joint was proposed. The user’s knee joint motion is detected via an IMU sensor, and a controller was provided to allow the exo-suit to track the human knee joint in synchrony. In this study, a conventional PID, SMC, and moving surface SMC controllers were designed, and the controllers’ performance were tested in real-time experiments. The maximum tracking errors in the PID, SMC, and MSMC controllers were 2.631 ∘ , 1.578 ∘ , and 1.289 ∘ , respectively, and the average tracking time errors were 0.10, 0.08, and 0.06 s, respectively. In addition, the designed and produced knee orthosis weighs 2.5 kg, making it one of the lightest and most compact designs for use by healthy individuals. The knee orthosis was made of steel, and thus it is very durable for use in all kinds of terrain conditions (military, industrial, etc.). Experimental findings about the proposed design and control method are analyzed in the results and discussion section.
Chapter
Full-text available
Training leads to increased neuronal excitability, decreased inhibition and different types of neuronal plasticity. Most studies focus on cortical plastic changes after cerebral lesions or in healthy humans. In this study, we investigate cortical excitability and plastic changes after a three month period of HAL® exoskeleton supported treadmill training in patients with chronic incomplete spinal cord injury by means of electrophysiological measurements and functional magnetic resonance imaging. Here we report preliminary results of four patients.
Article
Full-text available
Article
Full-text available
Wearable robots are a class of mechatronic systems intended to exchange energy with the environment and the human body in order to attain performance augmentation as well as for assistive, prosthetic or rehabilitative purposes. In this scenario a safe physical human-robot interaction assumes a crucial role both from hardware and software points of view. Whereas the conventional design methodology is effective in several robotics fields, issues arise in the case of wearable robots. The goal of the authors is to develop a novel wearable robots design methodology exploiting the concept of embodied intelligence. The paper starts from the description of what is a wearable robot and what are the design objectives to achieve. Then a state of the art of lower limbs wearable robots is reported. The adoption of a novel design methodology based on embodied intelligence is finally described and motivated. In conclusion, an example of the application of these new methods to a nonanthropomorphic wearable robot for gait restoration is reported.
Article
Full-text available
Systematic reviews and meta-analyses are essential to summarize evidence relating to efficacy and safety of health care interventions accurately and reliably. The clarity and transparency of these reports, however, is not optimal. Poor reporting of systematic reviews diminishes their value to clinicians, policy makers, and other users.Since the development of the QUOROM (QUality Of Reporting Of Meta-analysis) Statement--a reporting guideline published in 1999--there have been several conceptual, methodological, and practical advances regarding the conduct and reporting of systematic reviews and meta-analyses. Also, reviews of published systematic reviews have found that key information about these studies is often poorly reported. Realizing these issues, an international group that included experienced authors and methodologists developed PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) as an evolution of the original QUOROM guideline for systematic reviews and meta-analyses of evaluations of health care interventions.The PRISMA Statement consists of a 27-item checklist and a four-phase flow diagram. The checklist includes items deemed essential for transparent reporting of a systematic review. In this Explanation and Elaboration document, we explain the meaning and rationale for each checklist item. For each item, we include an example of good reporting and, where possible, references to relevant empirical studies and methodological literature. The PRISMA Statement, this document, and the associated Web site (http://www.prisma-statement.org/) should be helpful resources to improve reporting of systematic reviews and meta-analyses.
Article
We developed straight style transfer equipment for a person with disabled legs. It realizes travel in a standing state even on uneven ground, standing-up motion from a seated position, and ascending stairs. This equipment comprises three modules: a pair of telescopic crutches, a powered lower extremity orthosis, and a pair of mobile platforms. We detail the conceptual design of the equipment and the motion of each module. Cooperative operations using three modules are discussed through simulations. We verified travel in a standing state, including rotation, and standing-up motion from a chair through experiments using prototypes of telescopic crutches and mobile platforms.
Chapter
The International Classification of Functioning, Disability and Health (ICF), provides a framework for the description of health and health-related states and offers a biopsychosocial model of disability. The ICF was introduced by the World Health Organization in 1999 as a response to the conceptual and practical difficulties posed by its predecessor, the International Classification of Functioning, Disability and Health. The ICF lists body functions and structure, and activity and participation. The relationship between impairment, activity, and participation is not linear, and can be further moderated by contextual factors, including personal and environmental factors. There are a wide range of potential applications of the ICF.It has been adopted most widely within rehabilitation services to describe individual functioning, but can also be used at a service and national policy level to describe, monitor, and evaluate different activities.This chapter outlines the use of the ICF, considers its strengths, and highlights its function in a range of settings.
Chapter
Treadmill training after traumatic spinal cord injury is established as a therapy to improve walking capabilities in incomplete injured patients. In this study we investigate walking capabilities after a three month period of HAL® exoskeleton supported treadmill training in patients with chronic (>6 month) complete/incomplete (ASIA A – ASIA C) spinal cord injury. We monitored walking distance, walking speed and walking time with additional analysis of functional improvement by using the 10-m-walk test, the timed-up-and-go test and the WISCI II score in combination with the ASIA classification.
Article
The purpose of this methodological study was to evaluate the reliability, internal consistency, and sensitivity of a new self-report functional outcomes profile, the California Functional Evaluation 40 (CAFE 40). The CAFE 40 contains 3 sections: 1. Section 1 General health and activity level rated on an ordinal scale; 2. Section II: Independence- assistance needed rated on an ordinal scale; and 3. Section III: Endurance (specific listing of the patient reported amount of time that can be spent at a specific task such as walking, sitting, writing). In this study, Sections I and II were evaluated. Section III was rarely completed by the patients or normal subjects. 38 patients from 5 physical therapy outpatient clinics and 64 healthy adults participated in the reliability study by completing the functional profile twice within a 2 day interval. Test-retest reliability for the total profile score was high (r(lc) = 0.971). Internal consistency (the correlation between each item and the total section score reaching 0.5 or better) was acceptable for all but one item each in Section I and II (with the Spearman Rank Correlation Coefficient ranging from 0.408 to 0.82 in Section I and 0.463 to 0.748 in Section II). 45 patients participated in tile sensitivity study where the subjects completed the CAFE 40 at each physical therapy visit. A significant positive linear trend of improvement was noted (p < 0.001, Page Test). The last score was significantly higher than the first score for both Section I and II (p < 0.001 respectively as measured by Paired Wilcoxon Test). The results indicate that Sections I and II of the CAFE 40, have acceptable reliability, internal consistency, and sensitivity for measuring functional outcomes in healthy adults and patients receiving outpatient physical therapy. A revised profile is currently available for continued test development and broader application. Professionals in rehabilitation are encouraged to use this functional outcomes tool in their outpatient practices and send comments to the California Research Special Interest Group for continued improvement in the questionnaire.
Chapter
Rehabilitation robots have become an important tool in stroke rehabilitation. Compared to manual arm therapy, robot-supported arm therapy can be more intensive, of longer duration, and more repetitive. Therefore, robots have the potential to improve the rehabilitation process in stroke patients. In this chapter, the three-dimensional, multi-degree-of-freedom ARMin arm robot is presented. The device has an exoskeleton structure that enables the training of activities of daily living. Patient-responsive control strategies assist the patient only as much as needed and stimulate patient activity. This chapter covers the mechanical setup, the therapy modes, and the clinical evaluation of the ARMin robot. It concludes with an outlook on technical developments and about the technology transfer to industry.