Maciejasz2014 Article ASurveyOnRoboticDevicesForUppe

Maciejasz et al.
Journal of NeuroEngineering and Rehabilitation 2014, 11:3

http://www.jneuroengrehab.com/content/11/1/3 JNER JOURNAL OF NEUROENGINEERING
AND REHABILITATION
R EVIEW Open Access
A survey on robotic devices for upper limb

rehabilitation
Paweł Maciejasz1,2,3 , Jörg Eschweiler4* , Kurt Gerlach-Hahn5 , Arne Jansen-Troy4 and Steffen Leonhardt5
Abstract
The existing shortage of therapists and caregivers assisting physically disabled individuals at home is expected to
increase and become serious problem in the near future. The patient population needing physical rehabilitation of
the upper extremity is also constantly increasing. Robotic devices have the potential to address this problem as noted
by the results of recent research studies. However, the availability of these devices in clinical settings is limited, leaving
plenty of room for improvement. The purpose of this paper is to document a review of robotic devices for upper limb
rehabilitation including those in developing phase in order to provide a comprehensive reference about existing
solutions and facilitate the development of new and improved devices. In particular the following issues are discussed:
application field, target group, type of assistance, mechanical design, control strategy and clinical evaluation. This
paper also includes a comprehensive, tabulated comparison of technical solutions implemented in various systems.
Introduction these problems, prevailing research studies present a wide

An individual’s capacity to move is necessary to perform variety of devices specifically assisting physical rehabilita-
basic activities of daily living (ADL). Movement disorders tion. Robotic devices with the ability to perform repetitive
significantly reduce a patient’s quality of life. Disorders tasks on patients are among these technically advanced
of the upper extremities specifically limit the indepen- devices. In fact, robotic devices are already used in clinical
dence of affected subjects. Fortunately, there are vari- practice as well as clinical evaluation. However, consider-
ous approaches to restore the functionality of the upper ing the number of devices described in the literature, so
extremity, e.g., orthoses, functional electrical stimulation, far only a few of them have succeeded to target the subject
and physical therapy. Positive outcome of physical reha- group — for more details see Table 1. Furthermore, it
bilitation, in the case of neurologically based disorders, seems that the outcome of the use of devices already in
depends heavily on: onset, duration, intensity and task- clinical practice is not as positive as expected [3]. New
orientation of the training [1,2], as well as the patient’s solutions are being considered. Most of the literature
health condition, attention and effort [3]. Intense repe- reviews on robotic devices for upper extremity rehabilita-
titions of coordinated motor activities constitute a sig- tion (e.g. [6,7]) concentrate on devices that have already
nificant burden for the therapists assisting patients. In undergone clinical evaluation. Gopura and Kiguchi [8]
addition and due to economical reasons, the duration of compared the mechanical design of selected robotic
primary rehabilitation is getting shorter and shorter [4]. devices for upper extremity rehabilitation. However, no
These problems will probably exacerbate in the future other publication presents a summary of different robotic
as life expectancy continues to increase accompanied by solutions for upper extremity rehabilitation, including
the prevalence of both moderate and severe motor dis- those being in the development phase. An assessment
abilities in the elderly population [5] and consequently of different technical solutions would provide develop-
increasing their need of physical assistance. To counteract ers of robotic devices for upper limb rehabilitation an
evaluation of solutions that have already been consid-
*Correspondence: eschweiler@hia.rwth-aachen.de
ered, and thus learn from successes as well as shortfalls
4 Chair of Medical Engineering (mediTEC), Helmholtz-Institute for Biomedical from other research teams. Hence, a comparison of var-
Engineering, RWTH Aachen University, Pauwelsstraße 20, 52074 Aachen, ious robotic devices would facilitate the development of
Germany
Full list of author information is available at the end of the article
© 2014 Maciejasz et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative
Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.
Maciejasz et al. Journal of NeuroEngineering and Rehabilitation 2014, 11:3 Page 2 of 29
http://www.jneuroengrehab.com/content/11/1/3
Table 1 Robotic devices for upper limb rehabilitation

System DOF Supported Main Actuators Type; field of Stage of development;
name,references movements control inputs application additional information
Systems assisting shoulder movements
Kiguchi [114] 2 Shoulder – FE, AA sEMG DC motors (x2) Stationary system C0 study: 1 hs
(exoskeleton-based);
power assistance
Systems assisting elbow movements
Cheng [9] 1 Elbow – FE sEMG DC motor Stationary system; CI study: 5 stroke + 5 hs
physical therapy
Cozens [10] 1 Elbow – FE Joint angle Electric servo- Stationary system; CI study: 10 stroke + MS
motor physical therapy
Kiguchi [170] 1 Elbow – FE sEMG DC motor Stationary system C0 study: 2 hs
(exoskeleton-based);
physical therapy
MARIONET, Sulzer 1 Elbow – FE Joint angle AC servomotor Stationary system C0 study: 6 hs
[142] (SEA) (end-effector-based);
physical therapy
Mavroidis [11] 1 Elbow – FE Force/torque DC motor Portable orthosis Prototype
(continuous passive
motion device);
physical therapy
MEM-MRB, Oda [104] [1] [Elbow – flexion] Joint angular MRF brake Stationary system; C0 study: 1 hs
velocity, torque physical therapy
Myomo e100, Myomo, 1 Elbow – FE sEMG DC motor (x1) Portable orthosis; Commercial system (FDA
Inc.; Stein [172] physical therapy clearance); CI study: 8 cS
Ögce [171] 1 Elbow – FE sEMG DC step motor Wearable shoulder- CI study: 2 traumatic
elbow orthosis; brachial plexus injury
physical therapy
Pylatiuk [153] 1 Elbow – FE sEMG Hydraulic Wearable orthosis; First prototype
physical therapy
Rosen [169] 1 Elbow – FE sEMG DC motor (x1) Stationary system C0 study: 1 hs; predecessor
(exoskeleton-based); of CADEN-7
power assistance
Song [12] 1 Elbow – FE sEMG AC servo motor Stationary system CI studies: 8 cS [12], 7 cS
(end-effector-based); [13], 3 cS [14]
physical therapy
Vanderniepen [143] 1 Elbow – FE Joint angle Electric motors (x2) Wearable orthosis; Prototype
(SEA) orthopedic physical
therapy
Systems assisting forearm movements
Kung [15] 1 Forearm – PS Joint angle, AC servomotor Stationary system; CI study: 7 cS + 8 hs [16]
torque (1) physical therapy
Systems assisting wrist movements
ASSIST, Sasaki [146] 1 Wrist – flexion Joint angle Rotary-type Wearable orthosis; C0 study: 5 hs
pneumatic power assistance
actuators (x2)
Colombo [17] 1 Wrist – FE Torque Not specified Stationary system; CII study: 20(8) cS
physical therapy
Hu [18] 1 Wrist – FE sEMG Electric motor Stationary system CI study: 15 cS
(end-effector-based);
physical therapy
Loureiro [100] [1] [Wrist – FE] Hand motion MRF brake Wearable orthosis; CI study: 1 ET
(tremor) tremor suppression
PolyJbot, Song [175] 1 Wrist – FE sEMG, joint DC servomotor (x1) Stationary system; CII study: 27(15) cS [19]
angle and physical therapy
torque
Systems assisting finger(s) movements
Amadeo, tyromotion 5 Fingers (each) – End-point Electric motors Stationary system Commercial system; CI
GmbH FE position (end-effector-based); study: 7 aS [20]
and force physical therapy
Table 1 Robotic devices for upper limb rehabilitation (Continued)

Chen [21] 5 Independent Fingers DC linear motors Stationary system C0 study: 1 hs
linear movement positions and (x5) (end-effector-based);
of each finger forces, sEMG physical therapy
CyberGrasp, [5] [Resistive force to Joint angles DC motors (x5) Force-feedback Commercial system for
CyberGlove Systems each finger] (CyberGlove) glove; interactions other applications, used in
LLC; Turner [22] with virtual some clinical studies e.g.
environment [191,192]
Ertas [23] 1 Concurrent FE of Joint angles DC motor (x1) Finger exoskeleton C0 study: 4 hs
3 joints of a single (underactuated
finger mechanism); tendon
physical therapy
Fuxiang [24] 4 Index finger– FE Joint positions Linear stepping Modular-finger C0 study: 3 hs
(x3), AA and toques motors exoskeleton
(continuous passive
motion device);
physical therapy
Gloreha, Idrogenet srl 5 Independent Fingers Electric motors (x5) Portable (Gloreha Commercial system (CE
passive movement positions Lite)/Movable mark); CII study: 10(5) sS
of each finger (Gloreha [25], CI studies: 9 stroke +
Professional) (end- 3 other diseases [26], 4 cS
effector-based, [27]
cable-driven);
physical therapy
Hand of Hope, 5 Each finger sEMG DC linear motors Portable system Commercial system (CE
Rehab-Robotics separately - FE (x5) (orthosis); Mark), CI study: 8 cS
Comp. Ltd., Ho [28] physical therapy
HandCARE, Dovat 5 Independent Fingers DC motor (x1!) Stationary system CI study: 5 cS + 8 hs
[113] linear movement positions and (end-effector-
of each finger (1 forces based, cable-driven);
at a time) physical therapy
HEXORR, Schabowsky 2 Thumb – FE, Fingers DC motor (x1), AC Stationary system CI study: 5 cS + 9 hs
[29] other fingers positions and motor (x1) (end-effector-
together – FE forces based, cable-driven);
physical therapy
HIFE, Mali [183] 2 1 finger – FE End-point DC motors Haptic interface Prototype
position (end-effector-based);
physical therapy
InMotion HAND, 1 All fingers Not specified DC brushless motor Add-on module for Commercial system
previous name together – GR InMotion ARM;
InMotion 5.0, physical therapy
Interactive Motion
Tech., Inc.; Masia [165]
Kline [30] 1 All fingers Joint angles, Pneumatic Wearable glove; CI study: 1 stroke + hs (np)
together – sEMG physical therapy
extension
Lucas [147] 1 Index finger – sEMG Pneumatic (x2) Wearable orthosis; CI study: 1 SCI
flexion (passive grasp assistance
extension)
MR_CHIROD v.2, [1] [All fingers Finger position ERF brake Exercising device C0 study: hs (np);
Khanicheh [158] together – GR] and torque (handle-like); physical fMRI compatible
therapy
MRAGES, Winter [157] [5] [Fingers (each) -- FE] Finger positions MRF brakes (5) Force-feedback Prototype
and torques glove; physical
therapy
Mulas [31] 2 Thumb – FE, sEMG, pulleys DC servo motors Wearable orthosis; CI study: 1 sS
other fingers position (x2) physical therapy
together – FE
Nathan [167] 1 All fingers Hand-held FES Wearable orthosis CI study: 2 stroke + 1 hs
together – grasp trigger, index and (glove); physical
(passive release) thumb fingers therapy
joint angles
PowerGrip, Broaden 1 Thumb, index Switches DC motor (1) Wearable orthosis; Commercial system
Horizons, Inc. and middle finger or sEMG grasp assistance
together – GR

Reha-Digit, 1 4 fingers (except None DC motor Portable system Commercial system (CE
Reha-Stim; the thumb) (rotating handle); mark); CII study: 8(4) sS,
Hesse [32] together – FE physical therapy CI study: 1 cS
Rosati [144] 1 4 fingers (except Not DC motor (SEA) Wearable orthosis; Design
the thumb) selected yet physical therapy
together – FE
Rotella [33] 4 Index finger Not specified Electric motors Wearable orthosis; Design
flexion (x2) grasp assistance
(passive
extension),
thumb –
flexion, other
fingers together –
flexion
Rutgers Master II-ND, 4 Thumb, index, Actuator Pneumatic (x4) Force-feedback Research device;
Bouzit [184] middle, and ring translation and glove; interactions often used for
finger – FE inclination with virtual hand therapy
environment (e.g. [185-187])
Salford Hand 7 Index, middle, and Joint angles DC motors Wearable orthosis C0 study: hs (np)
Exoskeleton, ring finger – FE and end-point (exoskeleton);
Sarakoglou [34] (x2), thumb – FE force physical therapy
Tong [35] 10 Each finger – FE sEMG Electric linear Portable system CI study: 2 cS
(x2) motors (x10) (wearable orthosis);
physical therapy
TU Berlin Finger 4 1 finger – FE (x3), Joint angles DC motors (x4) Finger exoskeleton; C0 study: 1 hs
Exoskeleton, Wege [36] AA physical therapy
TU Berlin Hand 20 FE and AA of all Joint angles, DC motors Wearable orthosis Prototype
Exoskeleton, Fleischer major joints of end-point force, (exoskeleton);
[117] each finger sEMG physical therapy
Worsnopp [37] 3 Index finger – Joint angles DC brushless ser- Finger exoskeleton; Prototype
FE (x3) and torques vomotors (x6) physical therapy
Xing [38] 2 Thumb – FE, Position, force Pneumatic Wearable orthosis; C0 study: 3 hs
other fingers (PAMs) (x2) physical therapy
together – FE
Systems assisting shoulder and elbow movements
ACRE, Schoone [108] 5 Shoulder * elbow Joint angles Electrical motors Stationary system CI: 10 sS
(x5) (end-effector-based);
physical
therapy
ACT3D , Ellis [39] 3 Shoulder * elbow End-point DC brushed Stationary system CI study: 6 stroke
torque, motors (end-effector-based);
position and (HapticMaster) physical therapy
velocity and assessment of
(HapticMaster) therapy results
ARC-MIME, Lum [137] 1 + [2] Shoulder * elbow Forearm DC motor (x1), Stationary system An attempt to
(longitudinal position and magnetic particle (end-effector-based); commercialize; CI study: 4
movements of torque brakes (x2) physical cS; merges concepts from
the forearm) therapy MIME and ARM Guide
[forearm’s eleva-
tion and yaw]
ARM Guide, 1 + [2] Shoulder * elbow Forearm DC motor (x1), Stationary system CII study: 19(10) cS [40]; see
Reinkensmeyer [136] (longitudinal position and magnetic particle (end-effector-based); also: ARC-MIME
movements of torque brakes (x2) physical
the forearm) therapy
[forearm’s eleva-
tion and yaw]
BFIAMT, Chang [41] 2 Shoulder * elbow End point posi- DC servomotor Stationary system CI study: 20 cS
(bilateral longitu- tion and torque (x2), magnetic (end-effector-based); [41]
dinal movements particle brakes physical
of the forearms) (x2) therapy
BONES, Klein [118] 4 Shoulder – FE, AA, Joint angles, Pneumatic (x5) Stationary system Prototype; see also:
RT, elbow – FE cylinder (parallel robot + Supinator extender (SUE)
pressure exoskeleton-based
distal part); physical
therapy

Dampace, Stienen [4] [Shoulder – FE, Joint angles Hydraulic brake Stationary system CI study: stroke (np);
[154] AA, RT, elbow – and torques actuators (SEA) (exoskeleton-based); see also Limpact
FE] physical therapy
Freeman [163] 2 Shoulder * elbow Handle torque DC brusheless Stationary system C0 study: 18 hs
(in the plane) and position servomotors (x2), (end-effector-based);
FES physical therapy
InMotion ARM, 2 + [1] Shoulder * Joint positions, DC brushless Stationary system Commercial system,
previous name elbow (in the angular velocity motors (end-effector-based); CIII/CIV studies: 127(49) cS
InMotion 2.0, plane + gravity and torque physical therapy [203], CII studies: 56(30) aS
Interactive Motion compensation) [42], 30(10) aS [43] and
Tech., Inc.; based on: others
MIT Manus, Krebs
[107]
Ju [44] 2 Shoulder * elbow Handle torque AC motors (x2) Stationary system CI study: stroke (np)
(in the plane) and position (end-effector-based;
physical therapy
Kiguchi [45] 3 Shoulder – FE, AA, sEMG DC motors Wheelchair mounted C0 study: hs (np); see
elbow – FE system (exoskeleton- also: shoulder, elbow and
based); power shoulder-elbow-forearm
assistance orthoses developed by
Kiguchi and SUEFUL-7
Kobayashi [149] 4 Shoulder – FE, AA, Joint angle Pneumatic (PAMs) Wearable (but not C0 study: 5 hs
RT, elbow – FE (x10) portable) orthosis
(”muscle suit“); power
assistance
Limpact, Stienen [155] 4 Shoulder – FE, AA, Joint angles Rotational Stationary system Design; based on
RT, elbow – FE and torques hydroelastic (exoskeleton-based); Dampace
actuator (SEA) physical therapy
MariBot, Rosati [46] 5 Shoulder * elbow Motor positions DC frameless Stationary system Prototype; successor of
brushless motors (end-effector-based, NeReBot
cable-driven robot);
physical therapy
MEMOS, Micera [132] 2 Shoulder * elbow Torque and DC motors (x2) Stationary system CII study: 20(12) cS [17],
(in the plane) handle position (end-effector-based); CI study: 18 cS [47]
physical therapy
MIME, Burgar [120] 6 Shoulder * elbow Forearm DC brushed Stationary system CII studies: 27(13) cS
position, servomotors (end-effector-based); [48] and 30(24) sS [49],
orientation, (PUMA 560 robot) physical therapy CI study: 13 cS [50];
torque see also ARC-MIME
Moubarak [51] 4 Shoulder – FE, AA, Joint position, DC brushless Wheelchair- Prototype
RT, elbow – FE velocity and motors (x4) mounted system
torques (exoskeleton-based);
physical therapy
NeReBot, Rosati [111] 3 Shoulder * elbow Motor positions DC motors (x3) Stationary system CII studies: 24(12) sS
(end-effector-based, [111], 35 (17) aS [52], 21(11)
cable-driven robot); sS [53]; predecessor of
physical therapy MariBot
REHAROB, Toth [125] 12 Shoulder * elbow End-point Electrical motors Stationary system CII study: 22 (13) stroke +
torques (ABB IRB 140 and IRB (2 modified indus- 8(2) TBI [54], CI study:
1400H robots) trial robots); physical 6 cS + 2 sS + 4 hs [125]
therapy
Systems assisting forearm and wrist movements
Bi-Manu-Track, 1 Forearm – PS Not specified Not specified Stationary system Commercial system, CII
Reha-Stim; Hesse [55] * wrist – FE (end-effector-based); study: 44 (22) sA [56], CI
physical therapy study: 12 cS [55]
CRAMER, Spencer 3 Forearm – PS, Hand accelera- Digital servomotors Stationary system Prototype
[109] wrist – FE, AA tions (Nintendo (x4) (parallel robot);
Wii console) physical therapy
InMotion WRIST, 3 Forearm – PS, Joint angles DC brushless Stationary system, Commercial system
previous name wrist – FE * AA motors (x3) may be used as an
InMotion 3.0, add-on for InMotion
Interactive Motion ARM; physical
Tech., Inc.; Krebs [138] therapy

RiceWrist, Gupta [119] 4 Forearm – PS, Joint angles Frameless DC Wearable orthosis; Prototype; extension for
wrist – FE * AA and forces brushless motors physical therapy MIME, see also: MAHI
Supinator extender 2 Forearm – PS, Joint angles Pneumatic Wearable orthosis; CI study: 8 cS; extension for
(SUE), Allington [57] wrist – FE and forces physical therapy BONES and ArmeoSpring
Takaiwa [110] 3 Forearm – PS, Torque Pneumatic (x6) Stationary system Prototype
wrist – FE, AA (parallel robot);
physical therapy
W-EXOS, Gopura [174] 3 Forearm – PS, sEMG, hand DC motors (x3) Stationary system C0 study: 2hs; see also:
wrist – FE, AA force, forearm (exoskeleton-based); SUEFUL-7
torque power assistance
Systems assisting wrist and fingers movements
AMES, Cordo [58] 1 wrist and MCP Flexion/Extension Electric motor + Stationary system FDA clearance; CI study:
joints of 4 fingers torque, sEMG 2 vibrators (for (with desktop 20(11) cS; a modified
(coupled (optional) flexor and extensor mounted orthosis), version of the system
together) tendons) physical therapy (at is used for ankle
home) rehabilitation
Hand Mentor™, 1 Wrist and 4 fin- Wrist angle, Pneumatic (PAM) Wearable orthosis; Commercial system (FDA
Kinematic Muscles, gers (except the flexion torque (x1) physical therapy Class I Device); CII study:
Inc.; Koeneman [59] thumb) extension 21(11) sS [60], CI studies:
1 cS [61], 1 cS [62]
HWARD, Takahashi 3 Wrist – FE, Joint angles Pneumatic (x3) Stationary system CII study: 13(13) cS
[130] thumb – FE, and torques (with desktop
other fingers mounted orthosis);
together – FE physical therapy
My Scrivener, Obslap 3 Wrist * fingers End-point Electric motors Stationary system CI study: 18 children with
Reseach, LLC; Palsbo position and (Novint Falcon) (end-effector-based, weak handwriting skills
[190] torque (Novint using haptic device);
Falcon) fine motor hand
therapy
Systems assisting shoulder, elbow and forearm movements
ADLER, Johnson [63] 3 + {3} Shoulder * elbow End-point DC brushed motors Stationary system C0 study: 8 hs [64]
* forearm torque, position (HapticMaster) (end-effector-based);
and velocity physical therapy
(HapticMaster)
ARAMIS, Pignolo [65] 6x2 Shoulder – FE, AA, Joint angles DC brushed Stationary system CI study: 14 sS
RT, elbow – FE, and torques motors (x6 per (2 exoskeletons);
forearm – PS exoskeleton) physical therapy
Gentle/S, 3 + {3} Shoulder * elbow End-point DC brushed motors Stationary system CII study: 31(31) sS + cS
Amirabdollahian [121] * forearm torque, position (HapticMaster) (end-effector-based); [66]; predecessor of
and velocity physical therapy Gentle/G
(HapticMaster)
iPAM, Culmer [67] 6 Shoulder * elbow Joint torques Pneumatic Stationary system CI study: 16 cS
* forearm (2 robotic arms);
physical therapy
Kiguchi [68] 4 Shoulder – FE, AA, sEMG DC motors Wheelchair mounted C0 study: 1 hs; see also:
elbow – FE, system (exoskeleton- shoulder, elbow and
forearm – AA based); power assis- shoulder-elbow orthoses
tance developed by Kiguchi and
SUEFUL-7
L-Exos, Frisoli [197] 4 Shoulder – FE, AA, Joint angles Electric motors (x4) Stationary system CI study: 9 cS [69]
RT, elbow – FE (exoskeleton-based);
{forearm – PS} physical therapy
MGA, Carignan [70] 5 Shoulder – FE, AA, Joint torques DC brushless Stationary system Prototype
RT, VD, elbow – motors (x5) (exoskeleton-based);
FE, {forearm – PS} physical therapy
MULOS, Johnson [168] 5 Shoulder – FE, AA, Joystick (4 DOF) Electric motors (x5) Wheelchair- C0 study: 1 hs
RT, elbow – FE, mounted system
forearm – PS (exoskeleton-based);
power assistance
and physical therapy

NJIT-RAVR, Fluet [71] 3 + {3} Shoulder * End-point DC brushed motors Stationary system CI study: 8 CP
elbow * forearm torque, position (HapticMaster) (end-effector-based);
and velocity physical therapy
(HapticMaster) of children
RehabExos, Vertechy 4 Shoulder – FE, AA, Joint torques Custom-made Stationary system First prototype
[131] RT, elbow – FE frameless brushless (exoskeleton-based);
{forearm – PS} motor (x3), DC physical therapy
motor (x1)
Systems assisting shoulder, elbow and fingers movements
Pneu-WREX, 4 + {1} Shoulder – FE, AA, Joint angles, Pneumatic (x4) Stationary system CI study: 11 cS [72];
Wolbrecht [145] HD, elbow – FE, grasp force, (exoskeleton-based); see also: T-WREX and
{fingers – GR} cylinder physical therapy ArmeoSpring
pressure
T-WREX, Sanchez [106] {5} {Shoulder – FE, Joint angles, None Wheelchair mounted CII studies: 23(11) cS [73],
AA, RT, elbow – grasp force gravity balancing 28(14) cS [74], CI studies:
FE, fingers – GR} orthosis; physical 9cS + 5cS (2 studies) [75];
therapy see also: Pneu-WREX and
ArmeoSpring
Systems assisting elbow, forearm and wrist movements
Ding [179] 4 Elbow – FE, Joint angles Pneumatic (x8) Wearable (but not C0 study: 6 hs
forearm – PS, (a Motion portable) orthosis;
wrist – FE, AA Capture System power assistance for
is used) explicitly specified
muscles
MAHI, Gupta [76] 5 Elbow – FE, Joint angles Frameless DC Wearable orthosis Prototype; extension for
forearm – PS, brushless motors (force-feedback MIME; see also: RiceWrist
wrist – FE * AA exoskeleton);
physical therapy
WOTAS, Rocon [99] [3] [Elbow – FE, Angular DC motors (x3) Wearable orthosis; CI study: 10 mainly ET
forearm – PS, velocity, tremor suppression
wrist – FE] torques
Systems assisting forearm, wrist and fingers movements
Haptic Knob, 2 Forearm – Position, torque DC brushed motors Stationary system CI study: 3 cS
Lambercy [77] PS * wrist – FE, (x2) (2 parallelograms);
fingers – GR physical therapy
Hasegawa [98] 11 Forearm – PS, sEMG DC motors (x11) Wearable orthosis; C0 study: 1 hs
wrist – FE, AA, grasp assistance
thumb – FE (x2),
index finger –
FE (x3), other
fingers together –
FE (x3)
Kawasaki [178] 18 Forearm – PS, Joint angles of Servo motors (x22) Stationary system C0 study: 1 hs
wrist -- FE, thumb – healthy hand (exoskeleton-based);
FE (x3), AA, physical therapy
other fingers –
FE (x2), AA
Scherer [156] [1] [Forearm and Position, torque Magnetic particle Stationary system CI study: 2 stroke +
fingers twisting brake (end-effector-based, 1 MS
movements * rotating handle);
wrist – FE] physical therapy
Systems assisting shoulder, elbow, forearm and wrist movements
Braccio di Ferro, 2 Shoulder * Device joint AC brushless Stationary system CI studies: 10 cS + 4 hs [78],
Casadio [134] elbow * (forearm) * angles, end- servomotors (x2) (end-effector-based); 7 MS + 9 hs [79], 11 MS +
wrist (in the point force physical therapy 11 hs [80], 8 MS [81]
horizonatal or
vertical plane)
CADEN-7, Perry [97] 2x7 Shoulder – FE, sEMG, joint DC brushed motors Stationary system C0 study: 1 hs
AA, RT, elbow – angles, angular (2x7) (exoskeleton-based),
FE, forearm – PS, velocities and 2 robotic arms;
wrist – FE, AA forces/torques power assistance

Denève [82] 3 Shoulder * elbow Device joint AC brushless Stationary system Prototype
* (forearm) * wrist angles, end- motors (x3) (end-effector-based);
point force physical therapy
EMUL, Furusho [159] 3 Shoulder * End-point Electric motors + Stationary system CI study: 6 stroke;
elbow * (forearm) * position ERF clutches (end-effector-based); predecessor of PLEMO, see
wrist physical therapy also: Robotherapist
ESTEC exoskeleton, 9 Shoulder – FE, Joint angles Not selected yet Wearable system First prototype
Schiele [115] AA, RT, VD, HD, (exoskeleton-based);
elbow – FE, physical therapy
forearm – PS,
wrist – FE, AA
Furuhashi [83] 3 Shoulder * elbow * End-point DC motors (x3) Stationary system Prototype
(forearm) * wrist torque (end-effector-based);
physical therapy
Hybrid-PLEMO, Kikuchi 2 Shoulder * elbow * Device joint DC servomotors Stationary system Prototype; based on
[135] (forearm) * wrist angles, end- (x2) + ERF (end-effector-based); PLEMO
(in the adjustable point force clutches/brakes physical therapy
plane) (x4)
Lam [180] 2 Shoulder * elbow * End-point Not specified Stationary system C0 study: 8 hs
(forearm) * wrist position, abnor- (end-effector-based);
(in the plane) mal trunk posi- physical therapy
tion detection
Li [176] 5 Shoulder – FE, sEMG signals AC (x3) and DC (x2) Wearable system Prototype
AA, elbow – FE, from not servo motors (exoskeleton-based);
forearm – PS, affected arm physical therapy
wrist – FE
MACARM, Beer [112] 6 Shoulder * End-point posi- Electric motors (x8) Stationary system CI study: 5 cS
elbow * forearm * tion and force (end-effector-based,
wrist cable-driven robot);
physical therapy
Mathai [84] 3 Shoulder * End-point DC brushed motors Stationary system CI study: 4 cS
elbow * forearm * torque, position (HapticMaster) (end-effector-based);
wrist and velocity physical therapy
(HapticMaster)
MIME-RiceWrist, Gupta 10 Shoulder * See separate See separate Stationary system CI study: stroke (np)
[119] elbow * forearm * information for information for (robotic arm +
wrist MIME and MIME and RiceWrist orthosis); physical
RiceWrist system therapy
system
PLEMO, Kikuchi [105] [2] [Shoulder * elbow * Device joint ERF brakes Stationary system CI study: 6 stroke +
(forearm) * wrist] angles, end- (end-effector-based); 27 hs [85]; successor of
(in the adjustable point force physical therapy EMUL, predecessor of
plane) Hybrid-PLEMO
Robotherapist, 6 Shoulder * End-point Electric motors + Stationary system Prototype; see also:
Furusho [160] elbow * forearm * position ERF clutches (end-effector-based); EMUL
wrist physical therapy
RUPERT IV, 5 Shoulder – AA, Joint torques Pneumatic (PAMs) Wearable system CI study: 6 cS [86]
Balasubrama- nian RT, elbow – FE, and actuators (exoskeleton-based);
[151] forearm – PS, pressure physical therapy
wrist – FE
Salford Arm 7 Shoulder – FE, Joint positions Linear pneumatic Stationary system Prototype
Rehabilitation AA, RT, elbow – and torques actuators (PAMs) (exoskeleton-based);
Exoskeleton, FE, forearm – PS, (x14) physical therapy
Tsagarakis [148] wrist – FE, AA
Sophia-3, Rosati [87] 2 Shoulder * End-point posi- AC motors Stationary system First prototype; see also:
elbow * (forearm) * tion and force (end-effector-based, Sophia-4
wrist (in the planar cable-driven
plane) robot); physical
therapy
Sophia-4, Rosati [87] 2 Shoulder * End-point DC motors Stationary system Prototype; see also:
elbow * (forearm) * position and (end-effector-based, Sophia-3
wrist (in the force planar cable-driven
plane) robot); physical
therapy

SUEFUL-7, Gopura 7 Shoulder – FE, sEMG/joint DC servo motors Stationary system C0 study: 2 hs; shoulder-
[166] AA, RT, elbow – forces/torques (x7) (exoskeleton-based); elbow orthosis integrated
FE, forearm – PS, power assistance with W-EXOS system
wrist – FE, AA
Takahashi [182] 2 Shoulder * End point posi- Electric Stationary system CI study: 5 stroke +
elbow * tion servomotors (x2) (end-effector-based); 2 Guillain-Bare syndrome
(forearm) * wrist physical therapy
(in the plane)
Tanaka [88] 2 Shoulder * End-point force AC linear motor (x2) Stationary system C0 study: 6 hs
elbow * (forearm) * and position (end-effector-based);
wrist (in the physical therapy
plane)
UHD, Oblak [139] 2 3 configurations Torque and DC motors (x2), Stationary system CI study: 1 cS;
possible: 1) handle position (SEA) (end-effector-based); reconfigurable robot
shoulder * elbow, physical therapy
2) forearm – PS,
wrist – FE, 3)
forearm – PS,
wrist – AA
Umemura [152] 7 Shoulder – FE, Actuators Hydraulic Stationary system Prototype
AA, RT, elbow – pressure (end-effector-based);
FE, forearm – PS, physical therapy
wrist – FE, AA
UMH, Morales [127] 6 Shoulder * Joint torques Pneumatic Stationary system C0 study: hs (np)
elbow * forearm * (two robotic arms);
wrist physical therapy
Xiu-Feng [89] 2 Shoulder * Device joint AC servomotors Stationary system CI study: 30 stroke
elbow * angles, end- (x2) (end-effector-based);
(forearm) * wrist point force physical therapy
(in the plane)
Systems assisting shoulder, elbow, forearm, wrist and finger movements (whole arm)
ArmeoPower, Hocoma 6{ + 1} Shoulder – FE, Joint angles, DC motors (x6) Stationary system Commercial system; CI
AG; based on: ARMin AA, RT, elbow – grasp force (exoskeleton-based); studies: 3 cS (ARMin I)
III, Nef [90] FE, forearm – physical therapy [91], 4 cS (ARMin II) [92]
PS, wrist – FE,
{fingers – GR}
ArmeoSpring, Hocoma {7} {Shoulder – FE, Joint angles, None Stationary system Commercial system (CE
AG; based on: T-WREX, AA, RT, elbow – grasp force (exoskeleton-based); Mark, FDA clearance);
Sanchez [106] FE, forearm – physical therapy CI study: 10 MS [93]; see
PS, wrist – FE, also: T-WREX
fingers – GR}
ARMOR, Mayr [177] 8 Shoulder – FE, AA, Joint angles Electric motor Stationary master- CII study: 8(8) sS
RT, elbow – FE, of the master slave system
forearm – PS, hand (exoskeleton-based);
wrist – FE, physical therapy
thumb – FE,
other fingers
together – FE
Gentle/G, Loureiro 6{ + 3} Shoulder * End-point DC brushed Stationary system CII study: 4(4) sS [94];
[123] elbow (3 DOF, torque, position motors (robotic arm + based on Gentle/S
HapticMaster), and velocity (HapticMaster and orthosis); physical
{forearm – PS, (HapticMaster) grasp robot) therapy
wrist – FE, AA}, joint angels and
thumb – FE, end-point force
other fingers (grasp robot)
together – FE (x2)
(grasp robot)
HEnRiE, Mihelj [124] 4{ + 2} Shoulder * elbow End-point DC brushed Stationary system Prototype (with spring
(3 DOF, Haptic- torque, position motors (robotic arm + instead of an actuator in
Master), {wrist – and velocity (HapticMaster) orthosis); physical the hand part); C0 study: 1
FE, AA}, thumb, (HapticMaster) electric motors therapy hs; based on Gentle/S
middle and index joint angels and (grasping device)
finger together – end-point force
GR

IntelliArm, Ren [116] 8{ + 2} Shoulder – FE, AA, Joint angles Not specified Stationary system CI study: stroke (np)
RT, VD, {HD (x2)}, and torques (exoskeleton-
elbow – FE, fore- based); physical
arm – PS, wrist therapy
– FE, all fingers
together – GR
MUNDUS, Pedrocchi [3] + {2} +1 [Shoulder – FE, sEMG, button, elastic elements Modular CI study: 3 SCI + 2 MS
[101] AA, elbow – FE], eye-movement or DC brakes wheelchair-
optional: forearm – or BranComputer (shoulder- mounted system
PS, wrist – FE Interface; object elbow-wrist (exoskeleton-
(shoulder-elbow- labels – radio exoskeleton), FES based);
wrist exoskeleton), frequency (optional), DC movement
optional: all identification motor (optional assistance
fingers together – hand orthosis)
GR (hand orthosis)
ReoGo, Motorica 2 + {1} Shoulder * elbow; End-point posi- Electric motors Portable system Commercial system;
Medical Inc. also {* wrist} or tion (x4) (end-effector- CIII/CIV study: 60(np) sS
{fingers - FE} if based) with [198], CI studies: 14 cS [95],
special handle various handles; 10 sS [96]
used physical therapy
All the systems in the following table are grouped according to the joint movement they support. For the sake of convenience, we consider the shoulder complex, the
forearm and the hand (fingers) as single joints. Thus, we distinguish the following “joints”: shoulder, elbow, forearm, wrist and fingers. Devices assisting movements of
only one “joint” (starting from shoulder and ending with fingers) are described first followed by devices assisting movement of two, three and four joints (in that
order). The end of the table presents systems assisting movement of the whole arm.
For some systems it was difficult to classify them into a particular group. One of such cases includes the end-effector-based systems with a splint. A specific
classification to particular group may depend on the joints constrained in particular case by the splint. Furthermore, some devices allow for movements in some joints
only in a limited range.
In some cases the same system may appear multiple times in the table on various stages of development. We have accepted such occurrences only if, in our opinion,
the difference between two versions of the system justified considering them as two various systems. Otherwise, information included in the table includes only the
most recent version of the system available at the time of this publication.
System names are provided in italics. Whenever possible, the first column of the table provides the system name and reference (including the name of the first author)
to the publication in which the system is described. We only provide the appropriate reference for systems without a name. The names of commercial systems are
followed by their producer names. Appropriate information is provided following a semicolon for commercial systems based on systems being described in scientific
publication before commercialization. Except one case, i.e. ArmeoSpring based on T-WREX system, the description of the predecessors is not provided elsewhere in the
table because we found no significant differences between the predecessors and their commercial versions.
The last column contains information about the current stage of system development, clinical trials performed using the system and some additional information are
provided. If the system has undergone clinical evaluation, information about the category of the trial, number of participants enrolled and their condition, as well as
reference to the paper presenting results of the study is also provided. We distinguish four categories of the studies marked as C0, CI, CII, CIII/CIV. For a description, see
Table 7. Categories CII and CIII/CIV provide two numbers of subjects. The first number indicates the total number of participants enrolled in the study. The number in
parenthesis indicates number of participants undergoing therapy using the particular system. We made this distinction because there is often a control group
undergoing other form of therapy in the CII and CIII/CIV studies. If both numbers are equal, all participants underwent therapy using the specified system but other
parameter of the study varied between the groups (e.g. training intensity, device control strategy, or order in which various forms of therapy were applied). No
reference after the number and condition of participants indicates that the reference is the same as the one provided in the first column. Information about
predecessors or successors is also provided, if available. We use the following symbols and abbreviations:
- for degrees of freedom of the device (DOF) and supported movements (second and third column of the table respectively): [ ] - indicates passive (i.e. exerting only
resistive force) and { } - indicates not-actuated degrees of freedom or movements, otherwise active.
- for supported movements (third column): (joint name) - indicates that range of movements for that joint is limited to a very small range, AA – adduction/abduction,
FE – flexion/extension, GR – grasp and release, PS – pronation/supination, RT – internal/external rotation, HD - horizonatal displacement, VD - vertical displacement
(both in the shoulder girdle), MCP – metacarpophalangeal joint, * - indicates that the direction of the movement of the device does not correspond to the direction of
any of basic anatomical movements (e.g. pronation/supination, flexion/extension, rotation) but is a combination of many, (x number) - indicates that a few particular
movements are possible (e.g. flexion in a few joints of one finger), (in the plane) - indicates that the end effector of the device moves only in a specified plane; for the
explanation of anatomical terms of motion see Figure 2.
- for main control inputs and actuators (fourth and fifth column respectively): (commercial system name) - indicates that the particular commercial device (usually robot
or haptic interface) is incorporated in the described system and that the particular sensors or actuators are part of that commercial system.
- for main control (forth column): sEMG - surface electromyography.
- for actuators (fifth column): AC - alternating current, DC - direct current, ERF - electrorheological fluid based, FES - functional electrical stimulation, MRF -
magnetorheological fluid based, PAM - pneumatic artificial muscle, SEA - series elastic actuator, (x number) - number of particular actuators being used (provided only
if such an information was available).
- for clinical studies (last column): C0, CI, CII, CIII/CIV - category of the study: 0, I, II and III/IV, respectively (for category descriptions see the subsection Clinical studies of
the survey); subject condition: aS - acute stroke, CP - cerebral palsy, cS - chronic stroke, ET - essential tremor, hs - healthy subject(s), MS - multiple sclerosis, SCI - spinal
cord injury, sS - subacute stroke, TBI - traumatic brain injury; np - number of subjects is not provided.
new and improved devices for robotic upper limb reha- design of advance technology systems includes sensors,
bilitation. The aim of this paper is to summarize existing actuators, and control units; purely mechanical solutions
technical solutions for physical therapy of the upper limb. are excluded from this survey. Although the research team
The survey of robotic devices is comprised of advanced made an effort to identify as many systems as possi-
technology systems. As defined in this report, the ble, it is reasonable to acknowledge that many systems
are left unmentioned. Nevertheless, this documentation Throughout this report, the term “number of degrees
is intended to be a valuable source of information for of freedom (DOF)” describes the sum of all independent
engineers, scientists and physiotherapists working on new movements (i.e. displacements or rotations) that can be
solutions for physical rehabilitation. performed in all the joints of the device. The number of
DOF is defined in order to determine the exact position
The survey and orientation of all segments of the device. Also, some
Scope of the survey sections in this report are supplemented by an explana-
At the outset, the research team identified literature tion of the most important terminology for readers who
associated with the subject matter based on searches are not familiar with the technical vocabulary.
in PubMed, the Institute of Electrical and Electronics
Engineers (IEEE), Science Direct and Google Scholar Application field and target group
databases using various combinations of the following A description of the specific field of application for upper
keywords: upper extremity, arm, hand, rehabilitation, limb rehabilitation devices often determines solutions for
therapy, training, movement, motion, assistance, sup- which the device itself may be applied. Upper-extremity
port, robot, robotized, robotic, mechatronic, and motor- rehabilitation involves actions that stimulate patients’
ized. Additionally, referenced literature from the selected independence and quality of life. Two main application
publications was included in the survey as well. The fields of robotic devices stand out: support to perform
information obtained from this literature compendium is some ADLs (e.g. by power assistance or tremor suppres-
supplemented with the data acquired from professional sion) and providing physical training (therapy). Although
caregivers and manufacturers’ catalogs and websites, as there is a significant need for powered devices support-
well as direct communications with rehabilitation profes- ing basic ADL at home, there are only a few of such
sionals, manufacturers and patients. Over 120 systems are devices proposed so far (see sixth column in Table 1).
summarized and compared in Table 1; this tabulated sum- This is mainly due to technical and economical restric-
mary constitutes the reference for information provided in tions. Such devices should significantly improve the lives
subsequent sections. As previously mentioned, the scope of their users, otherwise patients become dissatisfied and
of this report is generally limited to the devices that sup- discontinue their use shortly after. They should be also
port or retrain movement or manipulation abilities of safe, easily to handle and inexpensive. Portability is also
disabled individuals. This survey excludes systems devel- often expected from devices assisting patients to perform
oped for movement assessment, occupational purposes or basic ADL; in such cases the amount of available energy
boosting physical abilities of healthy people. We however is limited by the capacity to store energy. Furthermore,
considered some academic, not yet specialized systems, if the device is supposed to support movements of mul-
supporting upper-extremity movements, especially if they tiple joints, the number of needed actuators increases as
have potential to be used for rehabilitation purposes (e.g. well as the weight of the device. Therefore, the number
CADEN-7 [97]). This survey also excludes devices that of portable actuated devices supporting upper extrem-
substitute movements of the disabled extremity but do ity movements is typically low. Instead, purely mechani-
not replace the movement itself (e.g. wheelchair mounted cal solutions are used for that purpose. A few examples
manipulators or autonomous robots). Although these of portable powered devices for upper extremity assis-
devices improve the patient’s quality of life, they differ tance used in daily living are PowerGrip system (Broaden
significantly from systems described in this survey and Horizons, Inc., USA) and a system proposed by Hasegawa,
constitute a separate category of devices. Some compa- et al. [98] (both for grasp assistance), as well as WOTAS
nies (e.g. CSMi Computer Sports Medicine, Inc.; Biodex orthosis [99] and a system proposed by Loureiro, et al.
Medical Systems, Inc.; BTE Technologies, Inc.) manufac- [100] (both for tremor suppression). However, portabil-
ture sensorized equipment for rehabilitation of various ity is not always necessary. Often, especially after a stroke
joints and muscles and whose principle of operation often or a spinal cord injury, disorders of the upper extrem-
resembles that of exercising devices found at fitness cen- ity are accompanied by lower extremity disabilities. These
ters. Those devices are used mainly to strengthen muscles scenarios are typically characterized by immobilized con-
and joints and provide some predefined resistance (e.g. ditions and require a wheelchair. Therefore, many sys-
isotonic, isometric or isokinetic exercises) or active force tems assisting upper limb movements are installed close
(e.g. continuous passive motion exercises). These devices to the patient (e.g. modular wheelchair-mounted system
also constitute a different category from the systems MUNDUS [101]).
included in this survey because their functions are per- Another group of the robotic devices used for rehabil-
formed along a predefined operation pathway. Although itation purposes, much bigger than the group of devices
difficult to clearly identify, the aforementioned were also supporting basic ADLs, constitute devices providing
excluded from this review. physical therapy. These may be designed for either
specialized therapeutic institutes or home-based condi- Table 2 Glossary of terms concerning type of assistance
tions. A vast majority of these devices may be used only Term Description
at therapeutic institutes since they require supervised Active device A device able to move limbs. Under such condition,
assistance from qualified personnel. Their price is often this device requires active actuators which may
prohibitive for personal use due to their complexity. The increase the weight. It may also apply to subjects
completely unable to move their limb.
patient demand for home-based therapy is expected to
increase. Along this context, the concept of the Gloreha Passive device A device unable to move limbs, but may resist the
movement when exerted in the wrong direction. This
system (Idrogenet srl) is provided in two versions: (1) a type of device may only be used for rehabilitation of
more complex and more adaptable professional version subjects able to move their limbs. It is usually lighter
intended for use at hospitals and rehabilitation centers and than active device since it possesses no actuators
other than brakes.
(2) a simplified low-cost version intended for patient use
at home. However, according to Dijkers, et al. [102], many Haptic device A device that interfaces with the user through the
sense of touch. In most cases it provides some
therapists may stop using devices if set-up takes more than amount of resistive force, often also some other
5 minutes. Thus new developed devices for physical train- sensation (e.g. vibration). It is sometimes also able to
ing should be intuitive, easy and fast to set-up and have a generate specific movements. However, the force it
generates is usually small. Haptic devices are
reasonable price. commonly used in rehabilitation settings with virtual
Stroke is the most common cause among diseases and environments.
injuries for upper limb movement disorders. It is esti- Coaching device A device that neither assists nor resists movement.
mated that by 2030, stroke will be the fourth leading However, it is able to track the movement and provide
cause of reduced disability-adjusted life-years (DALY) in feedback related to the performance of the subject.
As haptic devices, coaching devices are also
western countries. DALY takes into account years of life commonly used in rehabilitation settings with virtual
lost due to premature death as wells as years of life environments.
lived in less than full health [103]. Other causes include Active exercise An exercise in which subjects actively move their limb,
traumatic brain injury, spinal cord injury and injuries although some assistance of the device may be
to motoneurons, as well as certain neurological diseases provided. Such type of the exercise may be performed
using any of the above listed types of devices.
such as multiple sclerosis, cerebral palsy, Guillain-Barre
syndrome, essential tremor and Parkinson’s disease. Cur- Passive exercise An exercise in which the subject remains passive,
while a device moves the limb. This type of exercise
rently proposed robotic systems for upper limb rehabilita- requires an active device. Continuous passive motion
tion are typically tested on stroke patients. Only a fraction (CPM) training is an example of passive exercise with
of these systems are investigated on subjects suffering active devices.
from other diseases (see last column of Table 1).
Type of assistance the heavier actuators for active assistance. Devices using
The most important terminology introduced in this only resistive actuators include both devices for physical
section is explained in Table 2. Devices for upper limb therapy, e.g. MEM-MRB [104] and PLEMO [105], and sys-
rehabilitation may provide different types of motion assis- tems for tremor suppression, e.g. WOTAS [99] orthosis
tance: active, passive, haptic and coaching. Active devices and a system proposed by Loureiro, et al. [100].
provide active motion assistance and possess at least one Haptic devices constitute another group of systems
actuator, thus they are able to produce movement of the interacting with the user through the sense of touch.
upper-extremity. Most of the devices discussed in this sur- Haptic devices are similarly classified as either active or
vey are active (see Table 1). Such assistance of movements passive, depending on their type of actuator. In this report,
is required if patient is too weak to perform specific exer- haptic devices are independently categorized because
cises. However, even with active devices, an exercise is their main function is not to cause or resist movement
considered passive when a patient’s effort is not required. but rather to provide tactile sensation to the user. Other
For example, devices providing continuous passive motion non-actuated devices for upper limb rehabilitation do
exercises are active, but those exercises are categorized not generate any forces but provide different feedback.
as passive because the subject remains inactive while the These systems are labeled coaching devices through-
device actively moves the joint through a controlled range out this report. Because coaching devices are sensorized,
of motion. It is not necessary to apply active assistance they serve as input interface for interaction with thera-
to resist patient’s movement, to increase patient’s force or peutic games in virtual reality (VR) (e.g. T-WREX [106],
to ensure the patient is following the desired trajectory. ArmeoSpring from Hocoma AG) or for telerehabilitation
Instead, passive devices may be applied that are equipped (i.e. remotely supervised therapy). Coaching systems using
with actuators providing resistive force only (i.e. brakes). video-based motion recognition (e.g. Microsoft Kinect)
Such actuators consume less energy and are cheaper than would also belong to this category if it were not for their
lack of any mechanical part in contact with the patient. difference between the two categories is how the move-
Therefore, these systems will not be further discussed in ment is transferred from the device to the patient’s
this survey. upper extremity. End-effector-based devices contact the
Passive and non-actuated systems are less complex, patient’s limb only at its most distal part that is attached
safer and cheaper than their active counterparts. However, to patient’s upper extremity (i.e. end effector). Movements
they are often modified in the development process with of the end effector change the position of the upper limb
more active characteristics. Still, the main characteristic to which it is attached. However, segments of the upper
that identifies a non-actuated or passive device is the lack extremity create a mechanical chain. Thus, movements
of the ability to perform movement; they may be an option of the end effector also indirectly change the position of
for continuation of the rehabilitation process, rather than other segments of the patient’s body as well. Compared to
for training of people with significant movement disorders end effectors, exoskeleton-based devices have a mechani-
at an early stage of rehabilitation. cal structure that mirrors the skeletal structure of patient’s
limb. Therefore movement in the particular joint of the
Mechanical design device directly produces a movement of the specific joint
The most important terminology introduced in this of the limb.
section is explained in Table 3. When comparing the The advantage of the end-effector-based systems is
mechanical structure of robotic devices for movement their simpler structure and thus less complicated con-
rehabilitation often two categories of devices are control algorithms. However, it is difficult to isolate specific
sidered: end-effector-based and exoskeleton-based. The movements of a particular joint because these systems
produce complex movements. The manipulator allows up
to six unique movements (i.e. 3 rotations and 3 transla-
Table 3 Glossary of terms concerning mechanical design tions). Control of the movements of the patients upper
of robots for rehabilitation limb is possible only if the sum of possible anatomical
Term Description
movements of patient arm in all assisted joints is lim-
ited to six. Increasing the number of defined movements
End-effector based Contacts a subject’s limb only at its most distal
device part. It simplifies the structure of the device.
for the same position of the end-point of the manipulator
However, it may complicate the control of the results in redundant configurations of the patient’s arm,
limb position in cases with multiple possible thus inducing risk of injuries and complicated control
degrees of freedom.
algorithms.
Exoskeleton-based A device with a mechanical structure that mirrors The typical end-effector-based systems include serial
device the skeletal structure of the limb, i.e. each segment
of the limb associated with a joint movement is manipulators (e.g. MIT Manus [107] - Figure 1B, ACRE
attached to the corresponding segment of the [108]), parallel (e.g. CRAMER [109] and a system pro-
device. This design allows independent, posed by Takaiwa and Noritsugu [110], both for wrist
concurrent and precise control of movements in a
few limb joints. It is, however, more complex than
rehabilitation), and cable-driven robots (e.g. NeReBot
an end-effector based device. Orthoses restricting [111] - Figure 1C, MACARM [112]). The mechanical
or assisting movement in one or more joints may structure of HandCARE [113] may be also recognized as
be also considered exoskeleton-based devices.
the series of end-effector-based cable-driven robots, each
Planar robot A device, usually end-effector based, moving in a of which induce movement of one finger. In this system
specific plane. Design of planar robots, decreases
costs as well as the range of movements that a clutch system allows independent movement of each
may be exercised. Although this device performs finger using only one actuator.
movements in a plane, joints of the limb may still Application of the exoskeleton-based approach allows
move in a three-dimensional space.
for independent and concurrent control of particular
Back-drivability A property of mechanical design indicating that movement of patient’s arm in many joints, even if the
the patient is able to move the device, even when
the device is in passive state. It increases patient overall number of assisted movements is higher than six.
safety, because it does not constrain limb However, in order to avoid patient injury, it is necessary
movements and keeps patient’s limb in a to adjust lengths of particular segments of the manipu-
comfortable position.
lator to the lengths of the segments of the patient arm.
Modularity A property of a device indicating that optional Therefore setting-up such device for a particular patient,
parts may adapt it to a specific condition or simply
to perform additional exercises. especially if the device has many segments, may take a
significant amount of time. Furthermore, the position of
Reconfigurability A property of a device indicating that its
mechanical structure may be modified without the center of rotation of many joints of human body,
adding additional parts in order to adapt it to the especially of the shoulder complex [114], may change
condition of the subject or to perform other form significantly during movement. Special mechanisms are
of training.
necessary to ensure patient safety and comfort when an
Figure 1 Examples of mechanical structures of robotic devices for upper limb rehabilitation. A: ARM Guide - simple system using linear
bearing to modify orientation [136]; B: InMotion ARM - end-effector-based commercial system [133]; C: NeReBot - cable-driven robot, ©2007 IEEE.
Reprinted, with permission, from [111]; D: ArmeoPower - exoskeleton-based commercial system (courtesy of Hocoma AG).
exoskeleton-based robot assists the movements of these approaches. In the ArmeoSpring system (Hocoma AG)
joints [114]. For this reason, the mechanical and control for example, only the distal part – comprising the
algorithm complexity of such devices is usually signifi- elbow, forearm and wrist – is designed as an exoskele-
cantly higher than of the end-effector-based devices. The ton. Therefore the limb posture is statically fully deter-
complexity escalates as the number of DOF increases. mined (as in exoskeleton-based systems) and the shoulder
In case of systems for the rehabilitation of the whole joint is not constrained, allowing easy individual sys-
limb the number of DOF reaches nine (ESTEC exoskele- tem adaptation to different patients. A similar concept
ton [115]) or ten (IntelliArm [116]). Some systems for was applied in Biomimetic Orthosis for the Neuroreha-
fingers or hand rehabilitation have an even higher num- bilitation of the Elbow and Shoulder – BONES [118].
ber of DOF. Examples include the system proposed by In that case, a parallel robot consisting of passive slid-
Hasegawa, et al. with eleven DOF [98] and the hand ing rods pivoting with respect to a fixed frame pro-
exoskeleton developed at the Technical University (TU) vides shoulder movements. Such application of sliding
of Berlin with twenty DOF [117]. Even at such a high rods allows internal/external rotation of the arm without
number of DOF some of these devices still remain wear- any circular bearing element. The distal part allow-
able (i.e. the user is able to walk within a limited area ing for flexing/extending the elbow resembles the exos-
due to connections to power source and control unit, keleton structure. In the MIME-RiceWrist rehabilitation
e.g. ESTEC and hand exoskeleton developed at the TU system [119] the end-effector-based MIME [120] system
Berlin) or portable (i.e. the area within which the user for shoulder and elbow rehabilitation is integrated with
may walk is not limited, e.g. the system proposed by the parallel wrist mechanism used in the MAHI exoskele-
Hasegawa). ton (known as RiceWrist [119]).
Apart from purely exoskeleton- or end-effector-based Another example is the 6 DOF Gentle/S [121] system
devices, there are many systems combining a few allowing for relatively large reaching movements (three
actuated DOF of the end-effector-based commercial hap- [133] (Figure 1B). Designing the device as a planar robot
tic interface HapticMaster, Moog in the Netherlands BV reduces the range of movements that can be exercised for
[122]) and arbitrary positioning of the hand (connection particular joint. It also reduces the cost of the device. Fur-
mechanism with three passive DOF). The Gentle/S system thermore, when the working plane is well selected, the
was further supplemented with a three-active-DOF hand range of training motion may suffice in most of therapeu-
exoskeleton to allow grasp and release movements. This tic scenarios. Some of such planar devices allow changes
new nine DOF system is known as Gentle/G [123]. in the working space between horizontal and vertical
The HEnRiE [124] is a similar system based on the (Braccio di Ferro [134]) or even almost freely selecting the
Gentle/S system. In addition to the three active DOF of working plane (e.g. PLEMO and Hybrid-PLEMO [135]). It
HapticMaster, HEnRiE includes a connection mechanism further increases the range of possible exercise scenarios
with two passive DOF for positioning of the hand and while keeping the cost of the device at a minimum.
grasping device (two parallelogram mechanisms allowing In the ARM Guide [136] (Figure 1A) and ARC-MIME
parallel opening and closing of fingers attachments) with [137] systems, with which patients practice reaching
only one active DOF. movements, the working space is limited to linear move-
Some systems combine more than one robot at the ments because the forearm typically follows a straight-line
same time. This approach may be considered as the com- trajectory. However the orientation of the slide that assists
bination of end-effector-approach, where only the most forearm movements can be adjusted to reach multiple
distal parts of robots are attached to the patient’s upper workspace regions and fit different scenarios.
limb, with the exoskeleton-based approach, where move- Modularity and reconfigurability are concepts that
ments of few segments are directly controlled at the same could reduce therapy costs by adopting therapeutic
time. Use of two robots to control the movements of devices for various disabilities or stages in patient recov-
the limb may allow for mimicking the operations per- ery. However there are still only a few systems using these
formed by therapist using two hands. Examples of sys- concepts. For example, InMotion ARM robot (the com-
tems using two-robot-concept include REHAROB [125] mercial version of MIT Manus, previously called InMotion
(using two six-DOF manipulators), iPAM [126] and UMH 2.0) from Interactive Motion Technologies, Inc., may be
[127] (both having six DOF in total). Researchers at the extended by InMotion WRIST robot (previously InMotion
University of Twente, in Enschede, Netherlands, made an 3.0), developed at MIT [138] as a stand-alone system, and
attempt to use two HapticMaster systems to provide coor- InMotion HAND add-on module (previously InMotion
dinated bilateral arm training, but limitations in hardware 5.0) for grasp and release training. Another example of
and software caused the virtual exercise to behave differ- modular system is MUNDUS [101], consisting of various
ently to the real-life [128]. In some cases industrial robots modules that may be included depending on the patient
have also been used. The REHAROB uses IRB 140 and IRB condition, starting from muscle weakness to complete lost
1400H from ABB Ltd., while MIME [120] uses PUMA 560 of residual muscle function. For example as command
robot. In general, industrial robots reduce costs; however, input residual voluntary muscular activation, head/eye
such robots have significantly higher impedance than the motion, or brain signals may be used. However, this sys-
human upper limb and, according to Krebs, et al. [129], tem’s complexity might make commercialization of the
should not be in close physical contact with patients. device very difficult.
Therefore most of the robots used for the rehabilita- A very interesting solution was implemented in the Uni-
tion of the upper limb are designed with a low intrinsic versal Haptic Drive (UHD) [139]. It has only two DOF and,
impedance. Some of those devices are also back drivable depending on the chosen configuration, it can train either
(e.g. HWARD [130] and RehabExos [131]), meaning that shoulder and elbow during reaching tasks or forearm
the patient’s force is able to cause movement of those (flexion/extension) and wrist. For the latter setting option,
devices when they are in passive state. Back-drivability it is also possible to select a flexion/extension or prona-
further increases safety of the patient because the device tion/supination training for the wrist. See Figure 2 for an
does not constrain patient movements. It also allows for explanation of anatomical terms used for description of
using the device as an assessment tool to measure patient’s upper limb motion.
range of motion.
The majority of the devices presented in Table 1 allow Actuation and power transmission
movements in three dimensions; however there are also The most important terminology introduced in this
planar robots, i.e., systems allowing movements only on section is explained in Table 4. Traditionally, energy
a specified plane (e.g. MEMOS [132] and PLEMO [105]). to the actuators is provided in three forms: electric
Also the MIT Manus system initially allowed movements current, hydraulic fluid or pneumatic pressure. The selec-
only on one plane [107]. Subsequently, an anti-gravity tion of the energy source determines the type of actua-
module added possibility to perform vertical movements tors used in the system. Most of the devices for upper
Table 4 Glossary of terms concerning actuation of robots

for rehabilitation
Term Description
Electric actuators Actuators powered by electric current. They
are the most common because they easily
provide a relatively high power and are able
to store energy. There is a wide selection
of commercially available electric actuators;
however, some of them are heavy and/or
their impedance is too high for rehabilitation
settings.
Hydraulic actuators Actuators powered by hydraulic pressure
(usually oil). They are able to generate high
forces. Their system is relatively complex
considering the maintenance of pressurized
oil under pressure to prevent leakage.
Commercial hydraulic actuators are also
heavy, therefore, only specially designed
hydraulic actuators are used in rehabilitation
robotics.
Pneumatic actuators Actuators powered by compressed air. They
have lower impedance and weigh less than
electric actuators. Special compressors or
containers with compressed air are required
for power.
Pneumatic Artificial A special type of pneumatic actuator with
Muscle (PAM, McKibben an internal bladder surrounded by a braided
type actuator) mesh shell with flexible, but non-extensible
threads. Because of their specific design, an
actuator under pressure shortens, similarly to
the contracting muscle. It is relatively light
and exerts force in a single direction. It is
difficult to control because of its slow and
non-linear dynamic functions.
Figure 2 Main movements (degrees of freedom) of the upper
Series Elastic Actuator A generic name used for a mechanism with
extremity. 1: arm flexion/extension; 2: arm adduction/abduction; 3: (SEA) an elastic element placed in series with an
arm internal(medial)/external(lateral) rotation; 4: elbow flexion/ actuator. This solution is relatively often
extension; 5: forearm pronation/supination; 6: wrist flexion/extension; met in the design of rehabilitation robots. It
7: wrist adduction(ulnar deviation)/abduction(radial deviation); 8: decreases the inertia and intrinsic
hand grasp/release. impedance of the actuator to allow a more
accurate and stable force control and
increase patient safety.
Functional Electrical It is a technique that uses electrical current to
Stimulation (FES) activate nerves and contract their innervated
extremity rehabilitation use electric actuators but there muscles. It produces the movement of the
are also other systems with pneumatic and hydraulic actu- limb using natural actuators of the body.
ators. The electric actuators are most common because However, it is difficult to achieve precise and
repeatable movement using this technique
of their ease to provide and store electrical energy as and it may be painful for the patient.
well as their relatively higher power. Various types and
sizes of electrical motors and servomotors are currently
commercially available. Some authors (e.g. Caldwell and
Tsagarakis [140]) argue that electric actuators are too control [141], thus increasing the safety of the patient. The
heavy, compared to their pneumatic counterparts, and disadvantage of application with an elastic element is the
their impedance is too high to be used in rehabilita- lower functional bandwidth. Still, the field of rehabilita-
tion applications. However, the relatively high power- tion does not usually require high bandwidths. SEAs with
to-weight ratio of pneumatic actuators is achieved by electric motors are investigated in MARIONET [142] and
neglecting the weight of the power source. Adding an UHD [139] systems, as well as in systems proposed by
elastic element in series with the actuator may also miti- Vanderniepen, et al. (referred to as MACCEPA actuators)
gate the high impedance of electric motors. This concept [143] and Rosati, et al. [144].
lead to the development of the so called Series Elastic A few systems use pneumatic actuators. Pneumatic
Actuators (SEAs). SEAs decrease inertia and user inter- actuators are lighter and have lower inherent impedance
face impedance to provide an accurate and stable force than the electric counterparts. Because such actuators
require pneumatic pressure, the system is generally either et al. [156]. A few groups have also investigated the
stationary (e.g. Pneu-WREX [145]), its service area is lim- application of brakes incorporating magnetorheological
ited (e.g. ASSIST [146]) or the compressor is installed (MRF brakes) and electrorheological fluids (ERF brakes).
on the patient’s wheelchair (e.g. system proposed by These fluids change their rheological properties (i.e.
Lucas, et al. [147]). Special type of pneumatic actuators, viscosity) depending on the applied magnetic or elec-
called Pneumatic Artificial Muscles (PAMs), Pneumatic tric field, respectively. Thanks to those properties it
Muscle Actuators or McKibben type actuators are often is possible to achieve brakes with high-performance
used in rehabilitation robotics (e.g. Salford Arm Rehabili- (with rapid and repeatable brake torque) [105]. MRF
tation Exoskeleton [148] or system proposed by Kobayashi brakes are used in MRAGES [157] and MEM-MRB [104]
and Nozaki [149]). Such actuators consist of an internal systems. ERF brakes are used in PLEMO [105] and
bladder surrounded by braided mesh shell with flexible, MR_CHIROD v.2 [158] systems. The same group that
but non-extensible, threads. When the bladder is pres- developed the PLEMO also proposed ERF clutches to con-
surized, the actuator increases its diameter and shortens trol the force provided by an electric motor in active
according to its volume, thus providing tension at its ends systems. Such an actuation system was implemented in
[150]. Due to such physical configuration, PAMs’ weight is EMUL [159], Robotherapist [160] and Hybrid-PLEMO
generally light compared to other actuators, but also have [135] devices.
slow and non-linear dynamic response (especially large The natural actuators of body muscles can be used
PAMs), in consequence they are not practical for clini- instead of external actuators. For this purpose, an elec-
cal rehabilitation scenarios [131,151]. In addition, at least trical stimulation of the muscles leading to their contrac-
two actuators are necessary in order to provide antago- tion can be applied. This specific electrical stimulation
nistic movements due to the unidirectional contracting. is known as Functional Electrical Stimulation (FES).
The ASSIST system has a special type of PAM with rotary FES significantly reduces the weight of the device. From a
pneumatic actuators that allows bending movements therapeutic point of view, FES allows patients to exercise
[146]. muscles, improving muscle bulk and strength and pre-
A total of four systems using hydraulic actuators venting muscular atrophy [161]. It has been also shown
were identified in this survey. All four of them are not that FES, complemented by conventional physiotherapy,
standard and use actuators developed specially for that may enhance the rehabilitation outcome [162]. However,
purpose. Reasons to evade industrial hydraulic actu- FES may cause strong involuntary muscle contractions
ators include weight, impedance, fluid leakages and and can be painful for patient. Furthermore, it is dif-
difficulties to provide fluid. Large, noisy systems are ficult to control movements using FES because of the
usually necessary for that purpose. Mono-and bi-articular non-linear force characteristic of contracting muscles,
types of Hydraulic Bilateral Servo Actuators (HBSAs) muscles fatigue and dependency of the achieved contrac-
are used in the wheelchair-mounted exoskeleton pro- tion on the quality of the contact between stimulating
posed by Umemura, et al. [152]. Miniaturized and flex- electrodes and the body tissue. There are two commer-
ible fluidic actuators (FFA) were applied in the elbow cial systems using FES for upper limb rehabilitation: Ness
orthosis proposed by Pylatiuk, et al. [153]. Hydraulic H200 (Bioness, Inc., US) and NeuroMove (Zynex Medical,
SEAs are used in two other systems: the Dampace Inc., US). Two other systems combining FES with assis-
system [154] is equipped with powered hydraulic disk tive force were proposed by Freeman, et al. [163] and
brakes; the Limpact system [155], developed by the same Li, et al. [164].
group, uses an active rotational Hydro-Elastic Actuator When selecting actuators, it is also important to con-
(rHEA). sider their location, especially with exoskeleton-based
In passive systems, it is often desired to modify the mechanical structures. The actuators can be placed dis-
amount of resistance during the exercise. This modifi- tally, close to the joints on which they actuate (e.g.
cation increases the resistance when the patient departs ArmeoPower system, Figure 1D). This specification sim-
from the desired trajectory or to provide haptic feed- plifies the power transmission by using direct drives.
back for VR interactions. In existing systems, differ- However, it increases the weight of the distal part of the
ent solutions for provision of adjustable resistive force device and inertia and makes it more difficult to control
have been implemented. Powered hydraulic brakes, for the system. On the other hand, locating the actuators in
example, controlled by electromotors in a SEA are used the proximal part of the device, often in the part that
in Dampace system [154]. Magnetic particle brakes are remains constrained, reduces the weight and inertia of
used in ARM Guide [136] (Figure 1A), in its succes- the distal part. However, a power transmission mecha-
sor ARC-MIME [137] to resist other than longitudinal nism complicates the mechanical structure and may lead
movements of the forearm, and in the device for train- to difficulties with control due to friction. For exam-
ing of multi-finger twist motion proposed by Scherer, ple, the same group who developed InMotion HAND
system proposed an earlier prototype of the hand mod- by Rosen, et al. [169] and Kiguchi, et al. [170]) and portable
ule with eight active DOF and cable-driven mechanism systems (e.g. systems proposed by Ögce and Özyalçin
for power transmission. The friction in that mechanism [171] and Pylatiuk, et al. [153]). So far the most successful
and its level of complexity was too high for clinical sce- of those systems is the one DOF portable orthosis devel-
narios [165]. Nevertheless, there are systems, in which oped at the Massachusetts Institute of Technology (MIT),
power transmission using cables and gear drives was suc- Cambridge, US [172]. The system successfully sustained
cessfully applied, like for example CADEN-7 [97] and clinical trials, received FDA approval and was commer-
SUEFUL-7 [166]. cialized as Myomo e100 system (Myomo, Inc.) [173].
Examples of sEMG-controlled systems supporting move-
Control signals ments of other joints include those proposed by Kiguchi,
The most important terminology introduced in this et al. [114] for shoulder rehabilitation, W-EXOS [174] for
section is explained in Table 5. Various signals may be forearm and wrist rehabilitation, PolyJbot [175] for wrist
used as control input of the device. Switches are often rehabilitation, SUEFUL-7 [166] exoskeleton for whole
used to simplify design. Examples include the PowerGrip limb (excluding fingers) movement assistance, TU Berlin
system from Broaden Horizons, Inc., hand held triggers Hand Exoskeleton [117] for fingers rehabilitation, as well
(e.g. FES based system for grasp assistance proposed by as 11-DOF portable orthosis for grasp assistance pro-
Nathan, et al. [167]) and a joystick (e.g.MULOS [168]). posed by Hasegawa, et al. [98]. The sEMG signals from
Most of the systems having more complex control strate- the contralateral healthy limb have been also used to con-
gies use either kinematic, dynamic or a mix of both trol movements of the affected one (see system proposed
input signals (see Table 1 for a comparison). The type by Li, et al. [176]). The concept of using movements of
of the signal used as control input is partially determined the not affected limb to control motion of the affected
by the low-level control strategy and vice-versa. In some one has been also implemented in Bi-Manu-Track sys-
cases, signals provided by actuator encoders (concerning tem (Reha-Stim, Germany), ARMOR exoskeleton [177]
position or torque) may be directly used for control pur- and device proposed by Kawasaki, et al. [178]. Using the
poses. However, usually torque measured by the encoder other limb to control the affected one is especially useful
is a sum of the torque exerted by the user on device during rehabilitation after stroke. In cases of hemipare-
and internal torques in the device. Therefore, for better sis (or hemiplegia), often only one side of the body is
control of forces between patient and device, it is use- affected.
ful to apply additional sensors that will measure those In some systems also contact-less movement detec-
forces directly. tion methods have been used. For example, Ding, et al.
Some systems use surface electromyography (sEMG) [179] proposed a system to assist the load of arbitrary
as an input signal, which provides information about selected muscles using motion capture systems in order to
intention of the person to perform particular movement. calculate the actual muscle force.
Therefore it is possible to detect and support it. Most of
such systems support elbow movements, as sEMG signals Feedback to the user
from muscles controlling this joints (i.e. biceps brachii or Different types of feedback may be provided to the
triceps brachii) are relatively easily measured. Among pro- user, among them: visual, tactile, audio and in the form
posed solutions are both stationary (e.g. systems proposed of electrical stimulation. Some systems, for example
those proposed by Lam, et al. [180] and Nathan, et al.
[167], use vibrational stimulation of the muscle ten-
Table 5 Glossary of terms concerning input control signals dons to support their contraction. It was also suggested
of robots for rehabilitation that providing tactile feedback to flexor and extensor
Term Description surfaces of the skin at the appropriate location could
Dynamic signals Signals related to the torque or force exerted produce more naturalistic movements and improve clin-
by the subject on various joints of the device ical outcomes [3]. Some other systems combine other
(exoskeleton-based device) or at its end effector types of the feedback, for example system proposed by
(end-effector-based device).
Casellato, et al. [181] combines visual and haptic feed-
Kinematic signals Signals related to positions, orientations, velocities back to improve motor performance of children with
and accelerations of particular segments or joints
of the device or of the limb. dystonia.
A significant number of systems provide training in vir-
Trigger signal A signal initiating a specific action. In simple cases,
a switch or a button triggers the signal. In more tual reality (VR) scenarios. VR provides a much more
complex cases, a threshold value of some signal is interesting training surrounding to the patient, compared
specified to trigger the action (e.g. a sEMG value to the typically available conditions presented in therapeu-
corresponding to a level of muscle contraction).
tic units. Furthermore, it allows for fast modification of
training scenarios, increasing patient attention and moti- Table 6 Glossary of terms concerning control strategy of
vation to perform the exercise. Therefore it may also robots for rehabilitation
improve positive outcome of the therapy. It also adapts Term Description
the system for various patients in a very short time “High-level” control A control strategy with control algorithms
frame and restarts the task if the object was dropped strategy explicitly designed to induce motor plasticity.
or misplaced. Haptic devices are especially well suited Assistive control A “high-level” control strategy in which a device
for provision of therapy in VR because they provide provides the physical assistance to aid the patient
an impression of manipulating the virtual objects. Some in accomplishing an intended movement.
groups developed own versions of haptic systems. For Challenge-based A “high-level” control strategy in which a device
example Takahashi, et al. [182] proposed haptic device control challenges the patient to accomplish an intended
movement.
for arm rehabilitation, which can apply multiple types
of force including resistance, assistance, elasticity, vis- Haptic stimulation A “high-level” control strategy in which a robotic
device is used as a haptic interface to perform
cosity and friction. Other examples are: a two DOF activities in virtual reality environment.
Haptic Interface for Finger Exercise (HAFI), which pro-
Couching control A “high-level” control strategy in which a robotic
vides rehabilitation of only one finger at a time [183]; a device neither physically assists nor resists the
force reflecting glove, named MRAGES, using magneto- movement of the subject. It only quantifies and
rheological fluid [157]; MR_CHIROD v.2, a one DOF provides feedback (visual, acoustic or other)
concerning the performance of the subject during
grasp exercise device for functional magnetic resonance exercise.
imaging [158] and force-feedback glove Rutgers Mas-
“Low-level” control A control strategy considered in the implementa-
ter II-ND [184], developed at the Rutgers University strategy tion of the “high-level” control strategy in a device
(Piscataway, US) and used in hand therapy scenarios by appropriate control of the force, position,
(e.g. [185-187]). impedance or admittance.
Many groups have investigated application of a few Admittance control A “low-level” control strategy in which the force
exerted by the user is measured, and the device
of commercial haptic devices for rehabilitation of upper
generates the corresponding displacement.
extremity. Among such haptic interfaces are:
Impedance control A “low-level” control strategy in which the motion
- HapticMaster incorporated for example in Gentle/S of the limb is measured and the robot provides the
[121] (for other examples see Table 1), corresponding force feedback.
- in-parallel robots Phantom Omni and Premium (Geo- This terminology is mostly based on the one proposed by Marchal-Crespo and
magic, Inc., US) - used e.g. in experiments performed by Reinkensmeyer [193].
Casellato, et al. [181], Brewer, et al. [188], and Xydas and
Louca [189],
- parallel robot Falcon (Novint Technology, Inc., US) - “High-level” control strategies
used in My Scrivener system for hand writing training There is a myriad of “high-level” control strategies for
(Obslap Research LLC, US) [190], robotic movement training. This section briefly summa-
- force-feedback glove CyberGrasp (CyberGlove Systems rizes the classification of those strategies presented by
LLC,US) - used among others in therapeutic scenarios Marchal-Crespo and Reinkensmeyer [193]. They identify
investigated by Adamovich, et al. [191,192]. four categorizes of control strategies: (a) assistive con-
Because the entertainment industry have recently intro- trol, (b) challenge-based control, (c) haptic stimulation,
duced many new devices to capture motion of the healthy and (d) non-contacting coaching. Although some systems
people for interaction with VR-based games, it may be may fall into a few of these categorizes, this classification
expected that soon some of those devices will be also well illustrates main notions in the “high-level” control of
adapted for rehabilitation purposes, providing so called robotic devices for upper limb rehabilitation. Those con-
“serious games”. trol strategies in most cases correspond also to active,
passive, haptic and coaching types of motion assistance
Control strategy described before.
The most important terminology introduced in this The assistive control strategy makes tasks safer and
section is explained in Table 6. Following the exam- easier to accomplish, allowing more repetitions. There
ple of Marchal-Crespo and Reinkensmeyer [193] we are four types of assistive strategies: impedance-based,
will consider “high-” and “low-level” control strategies counterbalance-based, EMG-based and performance-
used by rehabilitation robots. “High-level” control algo- based adaptive control. In the impedance-based strat-
rithms are explicitly designed to provoke motor plasticity egy, the patient follows a particular trajectory. The
whereas “low-level” strategies control the force, position, device does not intervene as long as the patient fol-
impedance or admittance factors of the “high-level” con- lows this trajectory. However, as the patient leaves the
trol strategies. trajectory, the device produces a restoring force that
increases along with the deviation from the desired tra- “Low-level” control strategies
jectory. Often some margin of deviation from allowed Different “low-level” control strategies are combined to
trajectory is accepted before restoring force is pro- develop “high-level” rehabilitation strategies. Many “low-
vided. Counterbalance-based strategies provide a par- level” control strategies may be proposed during fol-
tial, passive or active weight counterbalance to a limb, lowing stages of development of a robotic rehabilitation
those making the exercises easier, as the amount of device. This report provides a short description of basic
force needed to move the limb against the gravity approaches and does not include a comprehensive com-
may be significantly reduced. EMG-based approach uses parison of “low-level” control strategies. General books on
the patient’s own sEMG signals to either trigger or control engineering provide a more detailed description,
proportionally control the assistance. Both of those in addition to articles referenced in Table 1.
approaches encourage patients’ effort. However, the trig- As the rehabilitation robots interact with human body,
gered method is more susceptible to slacking, as the it is necessary to consider the manipulator and patient
patient may learn to provide only the amount of force as a coupled mechanical system. The application of force
needed to trigger the assistance. Finally, the performance- or position control is not enough to ensure appropri-
based adaptive control strategies monitor the perfor- ate and safe dynamic interaction between human and
mance of the patient and adapt some aspects of the robot [195]. Other control strategies, such as impedance
assistance (e.g. force, stiffness, time, path) according to or admittance control are implemented in most of the
the current performance of the patient, as well as per- robots for upper limb rehabilitation. In the impedance
formance during particular number of preceding task control approach the motion of the limb is measured
repetitions. and the robot provides the corresponding force feed-
back, whereas in the admittance control approach the
Challenge-based control strategies fall into three force exerted by the user is measured, and the device
groups: resistive, amplifying error and constraint-induced. generates the corresponding displacement. The advan-
The resistive strategies resist the desired movements, tages and disadvantages of the impedance and admittance
those increasing the effort and attention of the patient. control systems are complementary [196]. In general,
The error amplifying strategies are based on the theory robots with impedance control have stable interaction
that faster improvements are achieved when error is but poor accuracy in free-space due to friction. This
increased [194]. Therefore they track the deviations from low accuracy can be improved using inner loop torque
the desired trajectories and either increase the observed sensors and low-friction joints or direct drives. Admit-
kinematic error or amplify its visual representation on tance control in contrast compensates the mass and
the screen. The constraint-induced robotic rehabilitation friction of the device and provides higher accuracy in
strategy, similarly to conventional constraint-induced non-contact tasks, but can be unstable during dynamic
therapy, promotes the use of the affected limb by cons- interactions. This problem is eliminated using SEAs.
training the not affected one. Devices using admittance control require also high trans-
mission ratios (e.g. harmonic drives) for precise motion
control [196]. In some cases both of theses approaches
Haptic stimulation strategies make use of haptic may be combined together. Impedance control strat-
devices described above, providing tactile sensation for egy has been implemented for example in MIT Manus
interactions with virtual reality objects. These strate- [107] (Figure 1B) and L-Exos exoskeleton [197], admit-
gies support training of basic ADLs in safe conditions tance control is found in MEMOS [132] and iPAM
and without long set-up. They provide alternate tasks in [126].
various environments, attracting attention and providing
conditions for implicit learning. Clinical evidence
Clinical studies
Non-contacting coaching strategy is applied in systems As previously discussed, there has been a significant effort
that do not contact participants but rather monitor their during last two decades to improve the design and con-
activity and provide instructions to the patient. Instruc- trol strategy of robotic rehabilitation devices. Yet, less has
tions indicate how to perform particular activities or what been done to prove the efficacy of such systems in rehabil-
should be improved. Since such devices do not contact itation settings. Although the results of clinical evaluation
the patient, they are not applicable for systems described of therapy applying robots are still sparse, the problem
herein. However, this category may be extended to is slowly being recognized. The focus in rehabilitation
include also some sensorized, but not-actuated exoskele- robotics is starting to move from technical laboratories to
tons, such as the gravity balancing orthosis T-WREX clinics. References to clinical trials in which robotic reha-
[106]. bilitation devices have been used are provided in the last
column of Table 1. The classification of clinical trials used that provides a therapy protocol rather than an end prod-
in this review is summarized in Table 7. uct [198], thus they are rather interested in responses to
From the developer and manufacturer’s point of view, questions concerning optimal training intensity, disorders
there may be at least three objectives in performing clinical for which such form of training may be beneficial, and
trials. The first one is to address regulatory requirements. whether robotic therapy should substitute or complement
The devices described in this review are considered med- other forms of therapy.
ical devices in most countries and as such the studies This survey includes a search into the US Clinical Trials
proving device efficacy and safety may be required before database (http://clinicaltrials.gov/) from October 2013
they are authorized for distribution. Although in some using a combination of keywords: robotic, robot, therapy
cases the exemption from providing the clinical data may and rehabilitation. This is an web-based database existing
be granted, e.g. if the device is recognized as low risk since 1997 and maintained by the US National Library of
(Class 1 device in the European Union and the USA) or Medicine at the National Institutes of Health. Under the
if equivalent device has been already approved for com- American Food and Drug Administration Amendments
mercialization, the clinical data may be required by health Act (FDAAA) of 2007 all the applicable clinical trials (what
insurance authorities in order to provide reimbursement. concerns category II and III/IV studies in our survey) per-
In this case the objective of the trial is to obtain a proof formed in the USA and starting after 2007 have to be
of clinical or financial benefit of the use of the device registered in this database. However, it includes also some
as compared to the existing modes of therapy. The third category I studies and many other studies performed in
objective of clinical trials is to provide the professional other countries. Results of this search identified 197 clin-
community with the clinical evidence of device’s efficacy. ical trials out of which 62 are relevant to this survey.
Although, the three objectives may seem to be similar, the The selected trials are divided into two categories. The
requirements are not the same, therefore when design- main objective of the first category is to proof the effi-
ing a clinical trial it should be considered if the obtained cacy or safety of the device, therefore there was either
results will allow to satisfy requirements of those three no control group or a control group was undergoing the
objectives. For the study design requirements to satisfy standard form of the therapy. The main objective of the
the marketing and reimbursement objectives, we refer the second category is to determine a more efficient form of
readers to the legal regulations in the country of interest. the therapy. In the latter category, the participants were
Whereas, for a review on the process to design a clinical assigned to groups undergoing similar forms of therapy,
train with sound scientific results we refer to Lo [198]. but at different intensities, using various devices or under-
From the clinical point of view, the objective of the going various forms of therapy in different order. A total
clinical study may be different than to validate a partic- of 31 studies aim at device safety or efficacy validation
ular device. For therapists the robotic device is a tool while 27 address better forms of therapy. A total of four
trials were excluded. The objective of these trials was to
validate other forms of therapy; devices described in this
Table 7 The classification of clinical trials of rehabilitation
review have only been used as a reference. As indicated in
robots used in this review
Figure 3, the number of participants enrolled in studies for
Term Description therapy improvement significantly increased during last
Category 0 Initial feasibility studies: Trials performed with low three years compared to the number of participants in
number of healthy volunteers, often using the prototype the device safety/efficacy validation studies. This suggests
of a device, in order to evaluate its safety and clinical
feasibility. that the objective of the studies changes from validating
forms of therapy to finding optimal applications methods.
Category I Pilot Consideration-of-Concept studies: Clinical trials
aimed at testing device safety, clinical feasibility and This survey identified a total of 21 devices out of the 58
potential benefit. They are performed in a small clinical studies. However, it was not possible to determine
population of subjects suffering from the target disease. the robotic device in 11 studies. Surprisingly, almost only
There is either no control group in the trial, or healthy
subjects are used as control. stroke survivors (54 studies) were enrolled. In the remain-
Category II Development-of-Concept studies: Clinical studies
ing four studies subjects with cerebral palsy, spinal cord
aiming at verification of device efficacy. Include a injury, traumatic brain injury and rotator cuff tear were
standardized description of the intervention, a control involved.
group, randomization and blinded outcome assessment.
Category III/IV Demonstration-of-Concept-Studies/Proof-of-Concept Outcomes of clinical studies
studies: Further evaluation of the device efficacy. Similar
to category II, however, usually these are multicentered
Many questions concerning effective robotic upper-limb
studies with high number of participants. rehabilitation still remain unanswered. One of the most
This classification is based on guidelines provided by Lo [198] and
important reasons is that the most effective interventions
supplemented by Category 0. to optimize neural plasticity are still not clear and it is not
Figure 3 Number of clinical studies and enrolled subjects depending on study objective and start year. Results based on a search in the
clinicaltrials.gov database in October 2013. Numbers on and above the bars indicate the number of studies in each category starting in the
particular year.
possible to implement them in rehabilitation robotics [7]. of the muscles) but appropriately timed activity of ago-
The other is that the results of the clinical controlled tri- nist and antagonist (coordination of the movements) that
als remains limited and those already available are difficult significantly improve the rehabilitation [3].
to compare with each other [7,193]. It is also questionable As previously stated, the objective of this report is
which measures should be used to evaluate the effects of not to review the results of clinical studies performed
therapy and which outcome should be compared: short- so far. A detailed review of clinical studies is referenced
term or long-term. Scales based on evaluation of abili- in other publications [7,198-202]. The most important
ties influencing the quality of life are often not objective results are still worth mentioning. Systematic review and
enough, since they rely on therapist expertise. meta-analysis of the trials performed in stroke patients
Although it is not possible to indicate the best con- suggest that robotic training improves motor impairment
trol strategy for the rehabilitation, there is already some and strength but do not improve ability to perform ADLs
evidence showing that some strategies are producing bet- [199,200]. The results of the first large randomized mul-
ter outcomes, whereas some can even decrease recov- ticenter study in which training with MIT-Manus robotic
ery time compared to possible non-robotic strategies system have been compared with intensive therapist-
[193]. provided therapy and usual care have revealed that there
is no significant difference in the outcomes of the two
Most accepted theories about robotic rehabilitation intensive forms of the therapy [203]. Thus the most
are clear: The goal of the rehabilitation training is not only important advantage of robotic systems is their abil-
to maximize the number of repetition but to maximize the ity to provide intensive repetitive training without over-
patients attention and effort as well [3]. The monotonous burdening therapists [204]. Another advantage is the
exercises provide worse retention of a skill compared with ability to provide more motivating training context, by
alternate training [7]. Adaptive therapy and assistance means of a computer gaming environment with quan-
as needed provide better results as fixed pattern ther- titative feedback to motivate practice [205]. Concerning
apy [193]. Robotic therapy can possibly decrease recovery cost-effectiveness of robotic rehabilitation, the results of
if it encourages slacking since the patient may decrease the previously mentioned multicenter trial have shown
effort and attention due to the use of adaptive algorithm that when the total cost of the therapy is compared, i.e.
[193]. Because learning is error based, faster improve- the cost of the therapy plus the cost of all the other health
ment may be achieved when error is increased [194]. care use, the costs of the two forms of the intensive ther-
Implicit learning, allowing patients to learn skills without apy (i.e. robot-assisted and therapist-provided) are similar
awareness, may result in greater learning effect [7]. Many [203]. However, the cost of technology is expected to
functional gains are more dependent on wrist and hand decrease, as opposed to the cost of human labor. There-
movements than on the mobility of shoulder and elbow fore cost-effective advantage toward robot-therapy may be
[7]. It is not the maximal voluntary contraction (strength expected [198].
Conclusions References
Due to population changes, shortage of professional ther- 1. Platz T: Evidenzbasierte Armrehabilitation: Eine systematische
Literaturübersicht [Evidence-based arm rehabilitation–a
apists, and the increasing scientific and technical poten- systematic review of the literature]. Nervenarzt 2003, 74(10):841–849.
tial, many research groups have proposed devices with [http://dx.doi.org/10.1007/s00115-003-1549-7]
the potential to facilitate the rehabilitation process. Many 2. Feys H, Weerdt WD, Verbeke G, Steck GC, Capiau C, Kiekens C, Dejaeger E,
Hoydonck GV, Vermeersch G, Cras P: Early and repetitive stimulation
devices for upper limb rehabilitation have already been of the arm can substantially improve the long-term outcome after
proposed. A vast majority of these proposed devices are stroke: a 5-year follow-up study of a randomized trial. Stroke 2004,
technically advanced and are designed for clinical set- 35(4):924–929. [http://dx.doi.org/10.1161/01.STR.0000121645.44752.f7]
3. Patton J, Small SL, Rymer WZ: Functional restoration for the stroke
tings. However, there is still significant need to improve survivor: informing the efforts of engineers. Top Stroke Rehabil 2008,
efficiency and reduce cost of home-based devices for ther- 15(6):521–541. [http://dx.doi.org/10.1310/tsr1506-521]
apy and ADLs assistance. The effectiveness of robotic 4. Richards L, Hanson C, Wellborn M, Sethi A: Driving motor recovery
after stroke. Top Stroke Rehabil 2008, 15(5):397–411.
over conventional therapy is arguable and the best ther- [http://dx.doi.org/10.1310/tsr1505-397]
apy strategy is still not clear. The situation may change 5. WHO: The global burden of disease: 2004 Update. World Health
soon, because more and more devices are being commer- Organization (WHO) 2008. [http://www.who.int/healthinfo/
global_burden_disease/GBD_report_2004update_full.pdf]
cialized and more scientific results will be available. It may 6. Riener R, Nef T, Colombo G: Robot-aided neurorehabilitation of the
encourage next groups to propose their own solutions. upper extremities. Med Biol Eng Comput 2005, 43:2–10.
Developing new devices and improving those already in 7. Brewer BR, McDowell SK, Worthen-Chaudhari LC: Poststroke upper
extremity rehabilitation: a review of robotic systems and clinical
the market will be easier, when taking advantage from the results. Top Stroke Rehabil 2007, 14(6):22–44.
already existing solutions. We hope that this survey will [http://dx.doi.org/10.1310/tsr1406-22]
help to navigate between those solutions and select best of 8. Gopura RARC, Kiguchi K: Mechanical designs of active upper-limb
exoskeleton robots: State-of-the-art and design difficulties. In Proc.
them, thus facilitating development of new and better sys- IEEE International Conference on Rehabilitation Robotics ICORR. Kyoto,
tems for robotic upper limb rehabilitation. We also hope Japan; 2009:178–187.
that it will be a valuable source of information for all the 9. Cheng HS, Ju MS, Lin CCK: Improving elbow torque output of stroke
patients with assistive torque controlled by EMG signals. J Biomech
professionals looking for a comprehensive reference.
Eng 2003, 125(6):881–886.
10. Cozens JA: Robotic assistance of an active upper limb exercise in
neurologically impaired patients. Rehabil Eng, IEEE Trans 1999,
Competing interests 7(2):254–256.
The authors declare that they have no competing interests. 11. Mavroidis C, Nikitczuk J, Weinberg B, Danaher G, Jensen K, Pelletier P,
Prugnarola J, Stuart R, Arango R, Leahey M, Pavone R, Provo A, Yasevac D:
Authors’ contributions Smart portable rehabilitation devices. J Neuroeng Rehabil 2005, 2:18.
PM performed systematical review of papers and drafted the manuscript. JE, [http://dx.doi.org/10.1186/1743-0003-2-18]
KGH, and AJ contributed to the review of various systems and revised the 12. Song R, yu Tong K, Hu X: Assistive control system using continuous
manuscript. SL contributed to the concept, paper structure and revised the myoelectric signal in robot-aided arm training for patients after
manuscript. All authors read and approved the manuscript. stroke. IEEE Trans Neural Syst Rehabil Eng 2008, 16(4):371–379. [http://dx.
doi.org/10.1109/TNSRE.2008.926707]
Acknowledgements 13. Hu X, Tong KY, Song R, Tsang VS, Leung PO, Li L: Variation of muscle
We would like to thank Mrs. Christine Azevedo-Coste, Ph.D. for her critical coactivation patterns in chronic stroke during robot-assisted
reading and revising the manuscript and Mr. Gabriel David Nevarez, P.E. for elbow training. Arch Phys Med Rehabil 2007, 88(8):1022–1029.
proofreading the English. [http://dx.doi.org/10.1016/j.apmr.2007.05.006]
The work of PM has been supported by the European Union in the framework 14. Song R, Tong KY, Hu XL, Tsang SF, Li L: The therapeutic effects of
of European Social Fund through the Warsaw University of Technology myoelectrically controlled robotic system for persons after
Development Program. He would also like to thank the Center for Advanced stroke–a pilot study. Conf Proc IEEE Eng Med Biol Soc 2006,
Studies of the Warsaw University of Technology for the support that he 1:4945–4948. [http://dx.doi.org/10.1109/IEMBS.2006.260186]
received during preparation of this work and the Chair of Medical Engineering 15. Kung PC, Ju MS, Lin CCK: Design of a forearm rehabilitation robot. In
(mediTEC), Helmholtz-Institute for Biomedical Engineering, RWTH Aachen Proc. IEEE 10th International Conference on Rehabilitation Robotics ICORR.
University for the possibility of performing the majority of the work on this Noordwijk, Netherlands; 2007:228–233.
manuscript during scientific stay in that institution. 16. Kung PC, Lin CCK, Ju MS, Chen SM: Time course of abnormal
synergies of stroke patients treated and assessed by a
neuro-rehabilitation robot. In Proc. IEEE International Conference on
Author details
1 DEMAR - LIRMM, INRIA, University of Montpellier 2, CNRS, Montpellier, 161 rue Rehabilitation Robotics ICORR. Kyoto, Japan; 2009:12–17.
17. Colombo R, Pisano F, Mazzone A, Delconte C, Micera S, Carrozza MC,
Ada, 34095 Montpellier, France. 2 Institute of Metrology and Biomedical
Dario P, Minuco G: Design strategies to improve patient motivation
Engineering, Warsaw University of Technology, ul. Św. A. Boboli 8, 02-525
during robot-aided rehabilitation. J Neuroeng Rehabil 2007, 4:3.
Warszawa, Poland. 3 Helmholtz-Institute for Biomedical Engineering, RWTH
[http://dx.doi.org/10.1186/1743-0003-4-3]
Aachen University, Pauwelsstraße 20, 52074 Aachen, Germany. 4 Chair of
18. Hu XL, Tong KY, Song R, Zheng XJ, Lui KH, Leung WWF, Ng S,
Medical Engineering (mediTEC), Helmholtz-Institute for Biomedical
Au-Yeung SSY: Quantitative evaluation of motor functional
Engineering, RWTH Aachen University, Pauwelsstraße 20, 52074 Aachen,
recovery process in chronic stroke patients during robot-assisted
Germany. 5 Philips Chair of Medical Information Technology (MedIT),
wrist training. J Electromyogr Kinesiol 2009, 19(4):639–650.
Helmholtz-Institute for Biomedical Engineering, RWTH Aachen University,
[http://dx.doi.org/10.1016/j.jelekin.2008.04.002]
Pauwelsstraße 20, 52074 Aachen, Germany.
19. Hu XL, Tong KY, Song R, Zheng XJ, Leung WWF: A comparison between
electromyography-driven robot and passive motion device on
Received: 25 October 2012 Accepted: 3 January 2014 wrist rehabilitation for chronic stroke. Neurorehabil Neural Repair
Published: 9 January 2014 2009, 23(8):837–846. [http://dx.doi.org/10.1177/1545968309338191]
20. Sale P, Lombardi V, Franceschini M: Hand robotics rehabilitation: 10th International Conference on Rehabilitation Robotics ICORR. Noordwijk,
feasibility and preliminary results of a robotic treatment in Netherlands; 2007:896–901.
patients with hemiparesis. Stroke Res Treat 2012, 2012:820931. 38. Xing K, Xu Q, He J, Wang Y, Liu Z, Huang X: A wearable device for
[http://dx.doi.org/10.1155/2012/820931] repetitive hand therapy. In Proc. 2nd IEEE RAS & EMBS International
21. Chen M, Ho SK, Zhou HF, Pang PMK, Hu XL, Ng DTW, Tong KY: Conference on Biomedical Robotics and Biomechatronics BioRob.
Interactive rehabilitation robot for hand function training. In Proc. Scottsdale, AZ; 2008:919–923.
IEEE International Conference on Rehabilitation Robotics ICORR. Kyoto, 39. Ellis MD, Sukal T, DeMott T, Dewald JPA: ACT 3D exercise targets
Japan; 2009:777–780. gravity-induced discoordination and improves reaching work area
22. Turner M, Gomez D, Tremblay M, Cutkosky M: Preliminary tests of an in individuals with stroke. In Proc. IEEE 10th International Conference on
arm-grounded haptic feedback device in telemanipulation. In Proc. Rehabilitation Robotics ICORR. Noordwijk, Netherlands; 2007:890–895.
of the ASME Dynamic Systems and Control Division. Volume 64. Anaheim, 40. Kahn LE, Lum PS, Rymer WZ, Reinkensmeyer DJ: Robot-assisted
CA; 1998:145–149. movement training for the stroke-impaired arm: Does it matter
23. Ertas IH, Hocaoglu E, Barkana DE, Patoglu V: Finger exoskeleton for what the robot does? J Rehabil Res Dev 2006, 43(5):619–630.
treatment of tendon injuries. In Proc. IEEE International Conference on 41. Chang JJ, Tung WL, Wu WL, Huang MH, Su FC: Effects of robot-aided
Rehabilitation Robotics ICORR. Kyoto, Japan; 2009:194–201. bilateral force-induced isokinetic arm training combined with
24. Fuxiang Z: An Embedded Control Platform of a Continuous Passive conventional rehabilitation on arm motor function in patients with
Motion Machine for Injured Fingers. In Rehabilitation Robotics. Edited chronic stroke. Arch Phys Med Rehabil 2007, 88(10):1332–1338.
by Kommu SS. Vienna, Austria: I-Tech Education Publishing; [http://dx.doi.org/10.1016/j.apmr.2007.07.016]
2007:579–606. 42. Volpe BT, Krebs HI, Hogan N, OTR LE, Diels C, Aisen M: A novel
approach to stroke rehabilitation: robot-aided sensorimotor
25. Vanoglio F, Luisa A, Garofali F, Mora C: Evaluation of the effectiveness
stimulation. Neurology 2000, 54(10):1938–1944.
of Gloreha (Hand Rehabilitation Glove) on hemiplegic patients.
43. Rabadi M, Galgano M, Lynch D, Akerman M, Lesser M, Volpe B: A pilot
Pilot study. In XIII Congress of Italian Society of Neurorehabilitation, 18-20
study of activity-based therapy in the arm motor recovery post
April. Italy: Bari; 2013.
stroke: a randomized controlled trial. Clin Rehabil 2008,
26. Parrinello I, Faletti S, Santus G: Use of a continuous passive motion
22(12):1071–1082. [http://dx.doi.org/10.1177/0269215508095358]
device for hand rehabilitation: clinical trial on neurological
44. Ju MS, Lin CCK, Lin DH, Hwang IS, Chen SM: A rehabilitation robot
patients. In 41 National Congress of Italian Society of Medicine and
with force-position hybrid fuzzy controller: hybrid fuzzy control of
Physical Rehabilitation, 14-16 October. Rome, Italy; 2013.
rehabilitation robot. IEEE Trans Neural Syst Rehabil Eng 2005,
27. Varalta V, Smania N, Geroin C, Fonte C, Gandolfi M, Picelli A, Munari D, 13(3):349–358. [http://dx.doi.org/10.1109/TNSRE.2005.847354]
Ianes P, Montemezzi G, La Marchina E: Effects of passive rehabilitation 45. Kiguchi K, Rahman MH, Sasaki M, Teramoto K: Development of a 3DOF
of the upper limb with robotic device Gloreha on visual-spatial and mobile exoskeleton robot for human upper-limb motion assist.
attentive exploration capacities of patients with stroke issues. In Robotics and Autonomous Systems 2008, 56(8):678–691. [http://www.
XIII Congress of Italian Society of Neurorehabilitation, 18-20 April. Bari, sciencedirect.com/science/article/B6V16-4R8MDRP-1/2/
Italy; 2013. 7d307e7bbef3e5958a6960e3da652723]
28. Ho NSK, Tong KY, Hu XL, Fung KL, Wei XJ, Rong W, Susanto EA: An 46. Rosati G, Gallina P, Masiero S, Rossi A: Design of a new 5 d.o.f.
EMG-driven exoskeleton hand robotic training device on chronic wire-based robot for rehabilitation. In Proc. 9th International
stroke subjects: task training system for stroke rehabilitation. IEEE Conference on Rehabilitation Robotics ICORR. Chicago, IL; 2005:430–433.
Int Conf Rehabil Robot; Boston, MA 2011, 2011:5975340. [http://dx.doi. 47. Colombo R, Sterpi I, Mazzone A, Delconte C, Minuco G, Pisano F:
org/10.1109/ICORR.2011.5975340] Measuring changes of movement dynamics during robot-aided
29. Schabowsky CN, Godfrey SB, Holley RJ, Lum PS, Development and pilot neurorehabilitation of stroke patients. IEEE Trans Neural Syst Rehabil
testing of HEXORR: hand EXOskeleton rehabilitation robot. J Eng 2010, 18:75–85. [http://dx.doi.org/10.1109/TNSRE.2009.2028831]
Neuroeng Rehabil 2010, 7:36. [http://dx.doi.org/10.1186/1743-0003-7-36] 48. Lum PS, Burgar CG, Shor PC, Majmundar M, der Loos, M V:
30. Kline T, Kamper D, Schmit B: Control system for pneumatically Robot-assisted movement training compared with conventional
controlled glove to assist in grasp activities. In Proc. 9th International therapy techniques for the rehabilitation of upper-limb motor
Conference on Rehabilitation Robotics ICORR. Chicago, IL; 2005:78–81. function after stroke. Arch Phys Med Rehabil 2002, 83(7):952–959.
31. Mulas M, Folgheraiter M, Gini G: An EMG-controlled exoskeleton for 49. Lum PS, Burgar CG, der Loos, M V, Shor PC, Majmundar M, Yap R: MIME
hand rehabilitation. In Proc. 9th International Conference on robotic device for upper-limb neurorehabilitation in subacute
Rehabilitation Robotics ICORR. Chicago, IL; 2005:371–374. stroke subjects: A follow-up study. J Rehabil Res Dev 2006,
32. Hesse S, Kuhlmann H, Wilk J, Tomelleri C, Kirker SGB: A new 43(5):631–642.
electromechanical trainer for sensorimotor rehabilitation of 50. Lum PS, Burgar CG, Shor PC: Evidence for improved muscle
paralysed fingers: a case series in chronic and acute stroke activation patterns after retraining of reaching movements with
patients. J Neuroeng Rehabil 2008, 5:21. [http://dx.doi.org/10.1186/ the MIME robotic system in subjects with post-stroke hemiparesis.
1743-0003-5-21] IEEE Trans Neural Syst Rehabil Eng 2004, 12(2):186–194. [http://dx.doi.org/
33. Rotella MF, Reuther KE, Hofmann CL, Hage EB, BuSha BF: An Orthotic 10.1109/TNSRE.2004.827225]
Hand-Assistive Exoskeleton for Actuated Pinch and Grasp. In 51. Moubarak S, Pham M, Pajdla T, Redarce T: Design Results of an Upper
Bioengineering Conference, IEEE 35th Annual Northeast. Boston, MA; Extremity Exoskeleton. In Proc. 4th European Conference of the
2009:1–2. [http://dx.doi.org/10.1109/NEBC.2009.4967693] International Federation for Medical and Biological Engineering. Antwerp,
34. Sarakoglou I, Tsagarakis NG, Caldwell DG: Occupational and physical Belgium; 2008.
therapy using a hand exoskeleton based exerciser. In Proc. IEEE/RSJ 52. Masiero S, Celia A, Rosati G, Armani M: Robotic-assisted rehabilitation
International Conference on Intelligent Robots and Systems (IROS). of the upper limb after acute stroke. Arch Phys Med Rehabil 2007,
Volume 3. Sendai, Japan; 2004:2973–2978. 88(2):142–149. [http://dx.doi.org/10.1016/j.apmr.2006.10.032]
35. Tong KY, Ho SK, Pang PK, Hu XL, Tam WK, Fung KL, Wei XJ, Chen PN, Chen 53. Masiero S, Armani M, Rosati G: Upper-limb robot-assisted therapy in
M: An intention driven hand functions task training robotic system. rehabilitation of acute stroke patients: focused review and results
In Conf Proc IEEE Eng Med Biol Soc. Volume 2010. Buenos Aires, Argentina; of new randomized controlled trial. J Rehabil Res Dev 2011,
2010:3406–3409. [http://dx.doi.org/10.1109/IEMBS.2010.5627930] 48(4):355–366.
36. Wege A, Hommel G: Development and control of a hand 54. Fazekas G, Horvath M, Troznai T, Toth A: Robot-mediated upper limb
exoskeleton for rehabilitation of hand injuries. In Intelligent Robots physiotherapy for patients with spastic hemiparesis: a preliminary
and Systems, (IROS 2005). 2005 IEEE/RSJ International Conference on. study. J Rehabil Med 2007, 39(7):580–582. [http://www.ingentaconnect.
Edmonton, Canada; 2005:3046–3051. com/content/mjl/sreh/2007/00000039/00000007/art00013]
37. Worsnopp TT, Peshkin MA, Colgate JE, Kamper DG: An Actuated Finger 55. Hesse S, Schulte-Tigges G, Konrad M, Bardeleben A, Werner C:
Exoskeleton for Hand Rehabilitation Following Stroke. In Proc. IEEE Robot-assisted arm trainer for the passive and active practice of
bilateral forearm and wrist movements in hemiparetic subjects. 71. Fluet GG, Qiu Q, Saleh S, Ramirez D, Adamovich S, Kelly D, Parikh H:
Arch Phys Med Rehabil 2003, 84(6):915–920. Robot-assisted virtual rehabilitation (NJIT-RAVR) system for
56. Hesse S, Werner C, Pohl M, Rueckriem S, Mehrholz J, Lingnau ML: children with upper extremity hemiplegia. In Virtual Rehabilitation
Computerized arm training improves the motor control of the International Conference. Haifa, Israel; 2009:189–192.
severely affected arm after stroke: a single-blinded randomized 72. Wolbrecht ET, Chan V, Reinkensmeyer DJ, Bobrow JE: Optimizing
trial in two centers. Stroke 2005, 36(9):1960–1966. [http://dx.doi.org/10. compliant, model-based robotic assistance to promote
1161/01.STR.0000177865.37334.ce] neurorehabilitation. IEEE Trans Neural Syst Rehabil Eng 2008,
57. Allington J, Spencer SJ, Klein J, Buell M, Reinkensmeyer DJ, Bobrow J: 16(3):286–297. [http://dx.doi.org/10.1109/TNSRE.2008.918389]
Supinator Extender (SUE): a pneumatically actuated robot for 73. Housman SJ, Le V, Rahman T, Sanchez RJ, Reinkensmeyer DJ: Arm-
forearm/wrist rehabilitation after stroke. Conf Proc IEEE Eng Med Biol Training with T-WREX After Chronic Stroke: Preliminary Results of a
Soc 2011, 2011:1579–1582. [http://dx.doi.org/10.1109/IEMBS.2011. Randomized Controlled Trial. In Proc. IEEE 10th International Conference
6090459] on Rehabilitation Robotics ICORR. Noordwijk, Netherlands; 2007:562–568.
58. Cordo P, Lutsep H, Cordo L, Wright WG, Cacciatore T, Skoss R: Assisted 74. Housman SJ, Scott KM, Reinkensmeyer DJ: A randomized controlled
movement with enhanced sensation (AMES): coupling motor and trial of gravity-supported, computer-enhanced arm exercise
sensory to remediate motor deficits in chronic stroke patients. for individuals with severe hemiparesis. Neurorehabil Neural
Neurorehabil Neural Repair 2009, 23:67–77. [http://dx.doi.org/10.1177/ Repair 2009, 23(5):505–514. [http://dx.doi.org/10.1177/
1545968308317437] 1545968308331148]
59. Koeneman EJ, Schultz RS, Wolf SL, Herring DE, Koeneman JB: 75. Sanchez RJ, Liu J, Rao S, Shah P, Smith R, Rahman T, Cramer SC, Bobrow JE,
A pneumatic muscle hand therapy device. Conf Proc IEEE Eng Med Biol Reinkensmeyer DJ: Automating arm movement training following
Soc 2004, 4:2711–2713. [http://dx.doi.org/10.1109/IEMBS.2004.1403777] severe stroke: functional exercises with quantitative feedback in a
gravity-reduced environment. IEEE Trans Neural Syst Rehabil Eng 2006,
60. Kutner NG, Zhang R, Butler AJ, Wolf SL, Alberts JL: Quality-of-life
14(3):378–389. [http://dx.doi.org/10.1109/TNSRE.2006.881553]
change associated with robotic-assisted therapy to improve hand
76. Gupta A, O’Malley M: Design of a haptic arm exoskeleton for training
motor function in patients with subacute stroke: a randomized
and rehabilitation. IEEE ASME Trans Mechatronics 2006, 11(3):280.
clinical trial. Phys Ther 2010, 90(4):493–504. [http://dx.doi.org/10.2522/
77. Lambercy O, Dovat L, Gassert R, Burdet E, Teo CL, Milner T: A haptic
ptj.20090160]
knob for rehabilitation of hand function. IEEE Trans Neural Syst Rehabil
61. Rosenstein L, Ridgel AL, Thota A, Samame B, Alberts JL: Effects of Eng 2007, 15(3):356–366. [http://dx.doi.org/10.1109/TNSRE.2007.903913]
combined robotic therapy and repetitive-task practice on 78. Casadio M, Giannoni P, Morasso P, Sanguineti V: A proof of concept
upper-extremity function in a patient with chronic stroke. Am J study for the integration of robot therapy with physiotherapy in
Occup Ther 2008, 62:28–35. the treatment of stroke patients. Clin Rehabil 2009, 23(3):217–228.
62. Frick EM, Alberts JL: Combined use of repetitive task practice [http://dx.doi.org/10.1177/0269215508096759]
and an assistive robotic device in a patient with subacute stroke. 79. Carpinella I, Cattaneo D, Abuarqub S, Ferrarin M: Robot-based
Phys Ther 2006, 86(10):1378–1386. [http://dx.doi.org/10.2522/ptj. rehabilitation of the upper limbs in multiple sclerosis: feasibility
20050149] and preliminary results. J Rehabil Med 2009, 41(12):966–970. [http://
63. Johnson M, Wisneski K, Anderson J, Nathan D, Smith R: Development of www.ingentaconnect.com/content/mjl/sreh/2009/00000041/
ADLER: The activities of daily living exercise robot. In 1st IEEE/ 00000012/art00004]
RAS-EMBS Int. Conf. Biomedical Robotics and Biomechatronics, BioRob. Pisa, 80. Casadio M, Sanguineti V, Solaro C, Morasso PG: A Haptic Robot Reveals
Italy; 2006:881–886. the Adaptation Capability of Individuals with Multiple Sclerosis. Int
64. Wisneski KJ, Johnson MJ: Quantifying kinematics of purposeful J Rob Res 2007, 26(11-12):1225–1233.
movements to real, imagined, or absent functional objects: 81. Vergaro E, Squeri V, Brichetto G, Casadio M, Morasso P, Solaro C,
implications for modelling trajectories for robot-assisted ADL Sanguineti V: Adaptive robot training for the treatment of
tasks. J Neuroeng Rehabil 2007, 4:7. [http://dx.doi.org/10.1186/1743- incoordination in Multiple Sclerosis. J Neuroeng Rehabil 2010, 7:37.
0003-4-7] [http://dx.doi.org/10.1186/1743-0003-7-37]
65. Pignolo L, Dolce G, Basta G, Lucca LF, Serra S, Sannita WG: Upper limb 82. Denève A, Moughamir S, Afilal L, Zaytoon J: Control system design of a
rehabilitation after stroke: ARAMIS a “robo-mechatronic” 3-DOF upper limbs rehabilitation robot. Comput Methods
innovative approach and prototype. In 4th IEEE RAS & EMBS Int. Conf. Programs Biomed 2008, 89(2):202–214. [http://dx.doi.org/10.1016/j.
Biomedical Robotics and Biomechatronics (BioRob). Rome, Italy; cmpb.2007.07.006]
2012:1410–1414. [http://ieeexplore.ieee.org/stamp/stamp.jsp? 83. Furuhashi Y, Nagasaki M, Aoki T, Morita Y, Ukai H, Matsui N:
arnumber=6290868] Development of rehabilitation support robot for personalized
66. Coote S, Murphy B, Harwin W, Stokes E: The effect of the GENTLE/s rehabilitation of upper limbs. In Proc. IEEE International Conference on
robot-mediated therapy system on arm function after stroke. Rehabilitation Robotics ICORR. Kyoto, Japan; 2009:787–792.
Clin Rehabil 2008, 22(5):395–405. [http://dx.doi.org/10.1177/ 84. Mathai A, Qiu Q: Incorporating Haptic Effects Into Three-Dimensional
0269215507085060] Virtual Environments to Train the Hemiparetic Upper Extremity; 2009.
67. Culmer PR, Jackson AE, Makower SG, Cozens JA, Levesley MC, [http://dx.doi.org/10.1109/TNSRE.2009.2028830]
Mon-Williams M, Bhakta B: A novel robotic system for quantifying 85. Ozawa T, Kikuchi T, Fukushima K, Akai H, Fukuda T, Tanida S, Fujikawa T,
arm kinematics and kinetics: description and evaluation in Kano S, Furusho J: Initial clinical tests for assessment models of
therapist-assisted passive arm movements post-stroke. J Neurosci synergy movements of stroke patients using PLEMO system with
Methods 2011, 197(2):259–269. [http://dx.doi.org/10.1016/j.jneumeth. sensor grip device. In Proc. IEEE International Conference on
2011.03.004] Rehabilitation Robotics ICORR. Kyoto, Japan; 2009:873–878.
68. Kiguchi K, Kose Y, Hayashi Y: Task-oriented perception-assist for an 86. Zhang H, Balasubramanian S, Wei R, Austin H, Buchanan S, Herman R,
upper-limb powerassist exoskeleton robot. In Proc. World He J: RUPERT closed loop control design. Conf Proc IEEE Eng Med Biol
Automation Congress (WAC). Kobe, Japan; 2010:1–6. [http://ieeexplore. Soc 2010, 2010:3686–3689. [http://dx.doi.org/10.1109/IEMBS.2010.
ieee.org/xpls/abs_all.jsp?arnumber=5665314] 5627647]
69. Frisoli A, Bergamasco M, Borelli L, Montagner A, Greco G, Procopio C, 87. Rosati G, Zanotto D, Secoli R, Rossi A: Design and control of two
Carboncini M, Rossi B: Robotic assisted rehabilitation in virtual planar cable-driven robots for upper-limb neurorehabilitation. In
reality with the L-EXOS. In Proc. of 7th ICDVRAT with ArtAbilitation. Maia, Proc. IEEE International Conference on Rehabilitation Robotics ICORR. Kyoto,
Portugal; 2008:253–260. Japan; 2009:560–565.
70. Carignan C, Tang J, Roderick S, Naylor M: A Configuration-Space 88. Tanaka Y, Ishii M, Tsuji T, Imamura N: Modeling and evaluation of
Approach to Controlling a Rehabilitation Arm Exoskeleton. In Proc. human motor skills in a virtual tennis task. Conf Proc IEEE Eng Med Biol
IEEE 10th International Conference on Rehabilitation Robotics ICORR. Soc; Vancouver, Canada 2008, 2008:4190–4193. [http://dx.doi.org/10.
Noordwijk, Netherlands; 2007:179–187. 1109/IEMBS.2008.4650133]
89. Xiu-Feng Z, Lin-Hong J, Li-Yun G: A novel robot neurorehabilitation 107. Krebs HI, Hogan N, Aisen ML, Volpe BT: Robot-aided
for upper limb motion. Conf Proc IEEE Eng Med Biol Soc; Shanghai, China neurorehabilitation. IEEE Trans Rehabil Eng 1998, 6:75–87.
2005, 5:5040–5043. [http://dx.doi.org/10.1109/IEMBS.2005.1615608] 108. Schoone M, van Os, P, Campagne A: Robot-mediated Active
90. Nef T, Guidali M, Klamroth-Marganska V, Riener R: ARMin - Exoskeleton Rehabilitation (ACRE) A user trial. In Proc. IEEE 10th International
Robot for Stroke Rehabilitation. In World Congress on Medical Physics Conference on Rehabilitation Robotics (ICORR). Noordwijk, Netherlands;
and Biomedical Engineering, September 7 - 12. Volume 25/IX of IFMBE 2007:477–481.
Proceedings. Edited by Dössel O, Schlegel WC. Munich, Germany: 109. Spencer SJ, Klein J, Minakata K, Le V, Bobrow JE, Reinkensmeyer DJ: A
Springer; 2009:127–130. low cost parallel robot and trajectory optimization method for
91. Nef T, Quinter G, Müller R, Riener: Effects of arm training with the wrist and forearm rehabilitation using the Wii. In Proc. 2nd IEEE RAS &
robotic device ARMin I in chronic stroke: three single cases. EMBS International Conference on Biomedical Robotics and
Neurodegener Dis 2009, 6(5-6):240–251. [http://dx.doi.org/10.1159/ Biomechatronics (BioRob). Scottsdale, AZ; 2008:869–874.
000262444] 110. Takaiwa M, Noritsugu T: Development of Wrist Rehabilitation
92. Staubli P, Nef T, Klamroth-Marganska V, Riener R: Effects of intensive Equipment Using Pneumatic Parallel Manipulator. In Proc. IEEE
arm training with the rehabilitation robot ARMin II in chronic International Conference on Robotics and Automation (ICRA). Barcelona,
stroke patients: four single-cases. J Neuroeng Rehabil 2009, 6:46. Spain; 2005:2302–2307. [http://dx.doi.org/10.1109/ROBOT.2005.
[http://dx.doi.org/10.1186/1743-0003-6-46] 1570456]
93. Gijbels D, Lamers I, Kerkhofs L, Alders G, Knippenberg E, Feys P: The 111. Rosati G, Gallina P, Masiero S: Design, implementation and clinical
Armeo Spring as training tool to improve upper limb functionality tests of a wire-based robot for neurorehabilitation. IEEE Trans Neural
in multiple sclerosis: a pilot study. J Neuroeng Rehabil 2011, 8:5. Syst Rehabil Eng 2007, 15(4):560–569. [http://dx.doi.org/10.1109/TNSRE.
[http://dx.doi.org/10.1186/1743-0003-8-5] 2007.908560]
94. Loureiro RCV, Lamperd B, Collin C, Harwin WS: Reach & grasp therapy: 112. Beer RF, Naujokas C, Bachrach B, Mayhew D: Development and
Effects of the Gentle/G System assessing sub-acute stroke evaluation of a gravity compensated training environment for
whole-arm rehabilitation. In Proc. IEEE International Conference on robotic rehabilitation of post-stroke reaching. In Proc. 2nd IEEE RAS &
Rehabilitation Robotics ICORR. Volume 2009. Kyoto; 2009:755–760. EMBS International Conference on Biomedical Robotics and
95. Bovolenta F, Sale P, Dall’Armi V, Clerici P, Franceschini M: Robot-aided Biomechatronics (BioRob). Scottsdale, AZ; 2008:205–210.
therapy for upper limbs in patients with stroke-related lesions. 113. Dovat L, Lambercy O, Gassert R, Maeder T, Milner T, Leong TC,
Brief report of a clinical experience. J Neuroeng Rehabil 2011, 8:18. Burdet E: HandCARE: a cable-actuated rehabilitation system to
96. Treger I, Faran S, Ring H: Robot-assisted therapy for neuromuscular train hand function after stroke. IEEE Trans Neural Syst Rehabil
training of sub-acute stroke patients. A feasibility study. Eur J Phys Eng 2008, 16(6):582–591. [http://dx.doi.org/10.1109/TNSRE.2008.
Rehabil Med 2008, 44(4):431–435. 2010347]
97. Perry JC, Rosen J, Burns S: Upper-limb powered exoskeleton design. 114. Kiguchi K, Iwami K, Yasuda M, Watanabe K, Fukuda T: An exoskeletal
Mechatronics, IEEE/ASME Trans 2007, 12(4):408–417. robot for human shoulder joint motion assist. Mechatronics,
IEEE/ASME Transactions on 2003, 8:125–135.
98. Hasegawa Y, Mikami Y, Watanabe K, Sankai Y: Five-fingered assistive
hand with mechanical compliance of human finger. In IEEE Int. Conf. 115. Schiele A, van der Helm, FCT: Kinematic design to improve
Robotics and Automation (ICRA). Pasadena, CA; 2008:718–724. ergonomics in human machine interaction. IEEE Trans Neural Syst
99. Rocon E, Belda-Lois JM, Ruiz AF, Manto M, Moreno JC, Pons JL: Design Rehabil Eng 2006, 14(4):456–469. [http://dx.doi.org/10.1109/TNSRE.2006.
and validation of a rehabilitation robotic exoskeleton for tremor 881565]
assessment and suppression. IEEE Trans Neural Syst Rehabil Eng 2007, 116. Ren Y, Park HS, Zhang LQ: Developing a whole-arm exoskeleton
15(3):367–378. [http://dx.doi.org/10.1109/TNSRE.2007.903917] robot with hand opening and closing mechanism for upper limb
100. Loureiro RCV, Belda-Lois JM, Lima ER, Pons JL, Sanchez-Lacuesta JJ, stroke rehabilitation. In Proc. IEEE International Conference on
Harwin WS: Upper limb tremor suppression in ADL via an orthosis Rehabilitation Robotics (ICORR). Kyoto, Japan; 2009:761–765.
incorporating a controllable double viscous beam actuator. In Proc. 117. Fleischer C, Kondak K, Wege A, Kossyk I: Research on Exoskeletons at
9th Int. Conf. on Rehabilitation Robotics ICORR. Chicago, IL; 2005:119–122. the TU Berlin. In Proceedings of German Workshop on Robotics, June 9-10.
101. Pedrocchi A, Ferrante S, Ambrosini E, Gandolla M, Casellato C, Schauer T, Braunschweig, Germany; 2009. [http://www.gwr09.org/download/
Klauer C, Pascual J, Vidaurre C, Gfoehler M, Reichenfelser W, Karner J, paper/gwr09_0043.pdf]
Micera S, Crema A, Molteni F, Rossini M, Palumbo G, Guanziroli E, 118. Klein J, Spencer S, Allington J, Bobrow JE, Reinkensmeyer DJ:
Jedlitschka A, Hack M, Bulgheroni M, D Amico E, Schenk P, Zwicker S, Optimization of a parallel shoulder mechanism to achieve a
Duschau-Wicke A, Miseikis J, Graber L, Ferrigno G: MUNDUS project: high-force, low-mass, robotic-arm exoskeleton. Robotics, IEEE Trans
MUltimodal Neuroprosthesis for daily upper limb support. 2010, 26(4):710–715.
J Neuroeng Rehabil 2013, 10:66. [http://dx.doi.org/10.1186/1743-0003- 119. Gupta A, O’Malley M, Patoglu V, Burgar C: Design, control and
10-66] performance of RiceWrist: a force feedback wrist exoskeleton for
102. Dijkers MP, deBear PC, Erlandson RF, Kristy K, Geer DM, Nichols A: rehabilitation and training. Int J Robot Res 2008, 27(2):233.
Patient and staff acceptance of robotic technology in occupational 120. Burgar CG, Lum PS, Shor PC, der Loos, H F M V: Development of robots
therapy: a pilot study. J Rehabil Res Dev 1991, 28(2):33–44. for rehabilitation therapy: the Palo Alto VA/Stanford experience.
103. Donnan GA, Fisher M, Macleod M, Davis SM: Stroke. Lancet 2008, 371 J Rehabil Res Dev 2000, 37(6):663–673. [http://www.rehab.research.va.
(9624):1612–1623. [http://dx.doi.org/10.1016/S0140-6736(08)60694-7] gov/jour/00/37/6/pdf/burgar.pdf]
104. Oda K, Isozumi S, Ohyama Y, Tamida K, Kikuchi T, Furusho J: 121. Amirabdollahian F, Loureiro R, Gradwell E, Collin C, Harwin W, Johnson
Development of isokinetic and iso-contractile exercise machine G: Multivariate analysis of the Fugl-Meyer outcome measures
MEM-MRB using MR brake. In Proc. IEEE Int. Conf. on Rehabilitation assessing the effectiveness of GENTLE/S robot-mediated stroke
Robotics (ICORR). Kyoto, Japan; 2009:6–11. therapy. J Neuroeng Rehabil 2007, 4:4. [http://dx.doi.org/10.1186/1743-
105. Kikuchi T, Xinghao H, Fukushima K, Oda K, Furusho J, Inoue A: 0003-4-4]
Quasi-3-DOF rehabilitation system for upper limbs: its 122. Van Der Linde R, Lammertse P: HapticMaster–a generic force
force-feedback mechanism and software for rehabilitation. In Proc. controlled robot for human interaction. Industrial Robot: Int J 2003,
IEEE 10th International Conference on Rehabilitation Robotics (ICORR). 30(6):515–524.
Noordwijk, Netherlands; 2007:24–27. 123. Loureiro RCV, Harwin WS: Reach & Grasp Therapy: Design and
106. Sanchez R, Reinkensmeyer D, Shah P, Liu J, Rao S, Smith R, Cramer S, Control of a 9-DOF Robotic Neuro-rehabilitation System. In Proc.
Rahman T, Bobrow J: Monitoring functional arm movement for IEEE 10th International Conference on Rehabilitation Robotics (ICORR).
home-based therapy after stroke. Conf Proc IEEE Eng Med Biol Soc; San Noordwijk, Netherlands; 2007:757–763.
Francisco, CA 2004, 7:4787–4790. [http://dx.doi.org/10.1109/IEMBS.2004. 124. Mihelj M, Podobnik J, Munih M: HEnRiE - Haptic environment for
1404325] reaching and grasping exercise. In Proc. 2nd IEEE RAS & EMBS
International Conference on Biomedical Robotics and Biomechatronics compliant actuation. In Proc. IEEE International Conference on
(BioRob). Scottsdale, AZ; 2008:907–912. Rehabilitation Robotics (ICORR). Kyoto, Japan; 2009:172–177.
125. Toth A, Fazekas G, Arz G, Jurak M, Horvath M, Passive robotic movement 144. Rosati G, Cenci S, Boschetti G, Zanotto D, Masiero S: Design of a
therapy of the spastic hemiparetic arm with REHAROB: report of the single-dof active hand orthosis for neurorehabilitation. In Proc. IEEE
first clinical test and the follow-up system improvement. In Proc. 9th International Conference on Rehabilitation Robotics (ICORR). Kyoto, Japan;
International Conference on Rehabilitation Robotics (ICORR). Chicago, IL; 2009:161–166.
2005:127–130. 145. Wolbrecht ET, Leavitt J, Reinkensmeyer DJ, Bobrow JE: Control of a
126. Culmer PR, Jackson AE, Makower S, Richardson R, Cozens JA, pneumatic orthosis for upper extremity stroke rehabilitation. Conf
Levesley MC, Bhakta BB: A Control Strategy for Upper Limb Robotic Proc IEEE Eng Med Biol Soc; New York, NY 2006, 1:2687–2693. [http://dx.
Rehabilitation With a Dual Robot System. Mechatronics, IEEE/ASME doi.org/10.1109/IEMBS.2006.259941]
Trans 2009, PP(99):1–11. 146. Sasaki D, Noritsugu T, Takaiwa M: Development of Active Support
127. Morales R, Badesa FJ, García-Aracil N, Sabater JM: Pneumatic robotic Splint driven by Pneumatic Soft Actuator (ASSIST). In Proc. IEEE
systems for upper limb rehabilitation. Med Biol Eng Comput 2011, International Conference on Robotics and Automation (ICRA). Barcelona,
49(10):1145–1156. [http://dx.doi.org/10.1007/s11517-011-0814-3] Spain; 2005:520–525.
128. Houtsma JA, Houten FJAMV: Virtual reality and a haptic master-slave 147. Lucas L, DiCicco M, Matsuoka Y: An EMG-controlled hand exoskeleton
set-up in post-stroke upper-limb rehabilitation. Proc Inst Mech Eng H for natural pinching. J Robot Mechatronics 2004, 16:482–488.
2006, 220(6):715–718. 148. Tsagarakis N, Caldwell D: Development and control of a
129. Krebs HI, Hogan N, Durfee W, Herr H: Rehabilitation robotics, orthotics and "soft-actuated" exoskeleton for use in physiotherapy and training.
prosthetics. In Textbook of Neural Repair and Rehabilitation(Selzer M, Autonomous Robots 2003, 15:21–33.
Clarke S, Cohen L, Duncan P, Gage F, eds.): Cambridge Univ Press; 2006. 149. Kobayashi H, Nozaki H: Development of muscle suit for supporting
manual worker. In IEEE/RSJ International Conference on Intelligent Robots
130. Takahashi CD, Der-Yeghiaian L, Le V, Motiwala RR, Cramer SC:
and Systems (IROS). San Diego, CA; 2007:1769–1774.
Robot-based hand motor therapy after stroke. Brain 2008,
150. Chou CP, Hannaford B: Measurement and modeling of McKibben
131(Pt 2):425–437. [http://dx.doi.org/10.1093/brain/awm311]
pneumatic artificial muscles. IEEE Trans Robot Automation 1996,
131. Vertechy R, Frisoli A, Dettori A, Solazzi M, Bergamasco M: Development 12:90–102.
of a new exoskeleton for upper limb rehabilitation. In Proc. IEEE 151. Balasubramanian S, Wei R, Perez M, Shepard B, Koeneman E, Koeneman
International Conference on Rehabilitation Robotics (ICORR). Kyoto, Japan; J, He J: RUPERT: An exoskeleton robot for assisting rehabilitation of
2009:188–193. arm functions. In Proc. Virtual Rehabilitation. Vancouver, Canada;
132. Micera S, Carrozza M, Guglielmelli E, Cappiello G, Zaccone F, Freschi C, 2008:163–167. [http://dx.doi.org/10.1109/ICVR.2008.4625154]
Colombo R, Mazzone A, Delconte C, Pisano F, et al.: A simple robotic 152. Umemura A, Saito Y, Fujisaki K: A study on power-assisted
system for neurorehabilitation. Autonomous Robots 2005, rehabilitation robot arms operated by patient with upper limb
19(3):271–284. disabilities. In Proc. IEEE International Conference on Rehabilitation
133. Krebs H, Ferraro M, Buerger S, Newbery M, Makiyama A, Sandmann M, Robotics (ICORR). Kyoto, Japan; 2009:451–456.
Lynch D, Volpe B, Hogan N: Rehabilitation robotics: pilot trial of a 153. Pylatiuk C, Kargov A, Gaiser I, Werner T, Schulz S, Bretthauer G: Design of
spatial extension for MIT-Manus. J Neuroeng Rehabil 2004, 1:5. [http:// a flexible fluidic actuation system for a hybrid elbow orthosis. In
dx.doi.org/10.1186/1743-0003-1-5] Proc. IEEE International Conference on Rehabilitation Robotics (ICORR).
134. Casadio M, Sanguineti V, Morasso PG, Arrichiello V: Braccio di Ferro: a Kyoto, Japan; 2009:167–171.
new haptic workstation for neuromotor rehabilitation. Technol 154. Stienen A, Hekman E, Prange G, Jannink M, Aalsma A, van der Helm F,
Health Care 2006, 14(3):123–142. van der Kooij H: Dampace: Design of an exoskeleton for
135. Kikuchi T, Ozawa T, Akai H, Furusho J: "Hybrid-PLEMO", rehabilitation force-coordination training in upper-extremity rehabilitation. J
system for upper limbs with active / passive force feedback, and its Med Devices 2009, 3:031003.
application for facilitation techniques. In Proc. IEEE International 155. Stienen AHA, Hekman EEG, ter Braak, H, Aalsma AMM, van der Helm FCT,
Conference on Rehabilitation Robotics (ICORR). Kyoto, Japan; van der Kooij H: Design of a rotational hydro-elastic actuator for an
2009:781–786. active upper-extremity rehabilitation exoskeleton. In Proc. 2nd IEEE
136. Reinkensmeyer DJ, Kahn LE, Averbuch M, McKenna-Cole A, Schmit BD, RAS & EMBS International Conference on Biomedical Robotics and
Rymer WZ: Understanding and treating arm movement impairment Biomechatronics (BioRob). Scottsdale, AZ; 2008:881–888.
after chronic brain injury: progress with the ARM guide. J Rehabil 156. Scherer R, Pradhan S, Dellon B, Kim D, Klatzky R, Matsuoka Y:
Res Dev 2000, 37(6):653–662. Characterization of multi-finger twist motion toward robotic
137. Lum P, Reinkensmeyer D, Mahoney R, Rymer WZ, Burgar C: Robotic rehabilitation. In Proc. IEEE International Conference on Rehabilitation
devices for movement therapy after stroke: current status and Robotics (ICORR). Kyoto, Japan; 2009:812–817.
challenges to clinical acceptance. Top Stroke Rehabil 2002, 8(4):40–53. 157. Winter SH, Bouzit M: Use of Magnetorheological fluid in a force
138. Krebs HI, Volpe BT, Williams D, Celestino J, Charles SK, Lynch D, Hogan N: feedback glove. IEEE Trans Neural Syst Rehabil Eng 2007, 15:2–8. [http://
Robot-aided neurorehabilitation: a robot for wrist rehabilitation. dx.doi.org/10.1109/TNSRE.2007.891401]
IEEE Trans Neural Syst Rehabil Eng 2007, 15(3):327–335. [http://dx.doi.org/ 158. Khanicheh A, Mintzopoulos D, Weinberg B, Tzika AA, Mavroidis C:
10.1109/TNSRE.2007.903899] MR_CHIROD v.2: magnetic resonance compatible smart hand
139. Oblak J, Cikajlo I, Matjacic Z: Universal Haptic Drive: A Robot for Arm rehabilitation device for brain imaging. Neural Syst Rehabil Eng IEEE
and Wrist Rehabilitation. IEEE Trans Neural Syst Rehabil Eng 2009. Trans 2008, 16:91–98.
[http://dx.doi.org/10.1109/TNSRE.2009.2034162] 159. Furusho J, Koyanagi K, Imada Y, Fujii Y, Nakanishi K, Domen K, Miyakoshi
140. Caldwell DC, Tsagarakis N: "Soft" Exoskeletons for Upper and Lower K, Ryu U, Takenaka S, Inoue A: A 3-D rehabilitation system for upper
Body Rehabilitation - Design, Control and Testing. Int J Humanoid limbs developed in a 5-year NEDO project and its clinical testing.
Robot 2007, 4:1–24. Chicago, IL; 2005:53–56.
141. Pratt GA, Williamson MM: Series elastic actuators. In Intelligent Robots 160. Furusho J, Kikuchi T, Oda K, Ohyama Y, Morita T, Shichi N, Jin Y, Inoue A:
and Systems 95. ’Human Robot Interaction and Cooperative Robots’, A 6-DOF Rehabilitation Support System for Upper Limbs including
Proceedings. 1995 IEEE/RSJ International Conference on. Volume 1. Wrists "Robotherapist" with Physical Therapy. In Proc. IEEE 10th
Pittsburgh, PA; 1995:399–406. International Conference on Rehabilitation Robotics (ICORR). Noordwijk,
142. Sulzer JS, Peshkin MA, Patton JL: Design of a Mobile, Inexpensive Netherlands; 2007:304–309.
Device for Upper Extremity Rehabilitation at Home. In Proc. IEEE 10th 161. Hamid S, Hayek R: Role of electrical stimulation for rehabilitation
International Conference on Rehabilitation Robotics (ICORR). Noordwijk, and regeneration after spinal cord injury: an overview. Eur Spine J
Netherlands; 2007:933–937. 2008, 17(9):1256–1269. [http://dx.doi.org/10.1007/s00586-008-0729-3]
143. Vanderniepen I, Van Ham R, Van Damme M, Versluys R, Lefeber D: 162. Takano Y, Haneda Y, Maeda T, Sakai Y, Matsuse H, Kawaguchi T,
Orthopaedic rehabilitation: A powered elbow orthosis using Tagawa Y, Shiba N: Increasing muscle strength and mass of thigh in
elderly people with the hybrid-training method of electrical preliminary design and evaluation results. J Neuroeng Rehabil 2008,
stimulation and volitional contraction. Tohoku J Exp Med 2010, 5:15. [http://dx.doi.org/10.1186/1743-0003-5-15]
221:77–85. 181. Casellato C, Pedrocchi A, Zorzi G, Vernisse L, Ferrigno G, Nardocci N:
163. Freeman CT, Hughes AM, Burridge JH, Chappell PH, Lewin PL, Rogers E: EMG-based visual-haptic biofeedback: a tool to improve motor
A robotic workstation for stroke rehabilitation of the upper control in children with primary dystonia. IEEE Trans Neural Syst
extremity using FES. Med Eng Phys 2009, 31(3):364–373. [http://dx.doi. Rehabil Eng 2013, 21(3):474–480. [http://dx.doi.org/10.1109/TNSRE.2012.
org/10.1016/j.medengphy.2008.05.008] 2222445]
164. Li R, Hu XL, Tong KY: Combined Electromyography(EMG)-driven 182. Takahashi Y, Terada T, Inoue K, Ito Y, Ikeda Y, Lee H, Komeda T: Haptic
system with functional electrical stimulation (FES) for poststroke Device System for Upper Limb Motor and Cognitive Function
rehabilitation. In Proc. 2nd IEEE RAS & EMBS International Conference on Rehabilitation: Grip Movement Comparison between Normal
Biomedical Robotics and Biomechatronics (BioRob). Scottsdale, AZ; Subjects and Stroke Patients. In Proc. IEEE 10th International Conference
2008:642–646. on Rehabilitation Robotics (ICORR). Noordwijk, Netherlands; 2007:
165. Masia L, Krebs HI, Cappa P, Hogan N: Design, Characterization, and 736–741.
Impedance Limits of a Hand Robot. In Proc. IEEE 10th International
183. Mali U, Munih M: HIFE-haptic interface for finger exercise.
Conference on Rehabilitation Robotics (ICORR). Noordwijk, Netherlands;
Mechatronics, IEEE/ASME Trans 2006, 11:93–102.
2007:1085–1089.
166. Gopura RARC, Kiguchi K, Li Y: SUEFUL-7: A 7DOF upper-limb 184. Bouzit M, Burdea G, Popescu G, Boian R: The Rutgers Master II-new
exoskeleton robot with muscle-model-oriented EMG-based design force-feedback glove. Mechatronics, IEEE/ASME Trans 2002,
control. In IEEE/RSJ International Conference on Intelligent Robots and 7(2):256–263.
Systems (IROS). St. Louis, MO; 2009:1126–1131. 185. Popescu VG, Burdea GC, Bouzit M, Hentz VR: A virtual-reality-based
167. Nathan DE, Johnson MJ, McGuire J: Feasibility of integrating FES telerehabilitation system with force feedback. IEEE Trans Inf Technol
grasp assistance with a task-oriented robot-assisted therapy Biomed 2000, 4:45–51.
environment: A case study. In Proc. 2nd IEEE RAS & EMBS International 186. Jack D, Boian R, Merians AS, Tremaine M, Burdea GC, Adamovich SV,
Conference on Biomedical Robotics and Biomechatronics (BioRob). Recce M, Poizner H: Virtual reality-enhanced stroke rehabilitation.
Scottsdale, AZ; 2008:807–812. Neural Syst Rehabil Eng, IEEE Trans 2001, 9(3):308–318.
168. Johnson GR, Carus DA, Parrini G, Marchese SS, Valeggi R: The design of 187. Adamovich SV, Merians AS, Boian R, Lewis JA, Tremaine M, Burdea GS,
a five-degree-of-freedom powered orthosis for the upper limb. Recce M, Poizner H: A virtual reality-based exercise system for hand
Proc Inst Mech Eng H 2001, 215(3):275–284. rehabilitation post-stroke. Presence: Teleoperators & Virtual Environ
169. Rosen J, Brand M, Fuchs MB, Arcan M: A myosignal-based powered 2005, 14(2):161–174.
exoskeleton system. IEEE Trans Syst, Man and Cybernet, Part A 2001, 188. Brewer BR, Klatzky R, Matsuoka Y: Visual feedback distortion in a
31(3):210–222. robotic environment for hand rehabilitation. Brain Res Bull 2008,
170. Kiguchi K, Esaki R, Tsuruta T, Watanabe K, Fukuda T: An exoskeleton 75(6):804–813. [http://dx.doi.org/10.1016/j.brainresbull.2008.01.006]
system for elbow joint motion rehabilitation. In Proc. IEEE/ASME 189. Xydas EG, Louca LS: Design and Development of a Haptic Peg-Board
International Conference on Advanced Intelligent Mechatronics (AIM). Exercise for the Rehabilitation of People with Multiple Sclerosis. In
Volume 2. Port Island, Japan; 2003:1228–1233. vol.2. Proc. IEEE 10th International Conference on Rehabilitation Robotics (ICORR).
171. Ögce F, Özyalçin H: Case study: A myoelectrically controlled Noordwijk, Netherlands; 2007:906–916.
shoulder-elbow orthosis for unrecovered brachial plexus injury. 190. Palsbo SE, Hood-Szivek P: Effect of robotic-assisted
Prosthet Orthot Int 2000, 24(3):252–255. three-dimensional repetitive motion to improve hand motor
172. Stein J, Narendran K, McBean J, Krebs K, Hughes R: function and control in children with handwriting deficits: a
Electromyography-controlled exoskeletal upper-limb-powered nonrandomized phase 2 device trial. Am J Occup Ther 2012,
orthosis for exercise training after stroke. Am J Phys Med Rehabil 2007, 66(6):682–690. [http://dx.doi.org/10.5014/ajot.2012.004556]
86(4):255–261. [http://dx.doi.org/10.1097/PHM.0b013e3180383cc5]
191. Adamovich S, Fluet GG, Merians AS, Mathai A, Qiu Q: Recovery of hand
173. Stein J: e100 NeuroRobotic system. Expert Rev Med Devices 2009,
function in virtual reality: Training hemiparetic hand and arm
6:15–19. [http://dx.doi.org/10.1586/17434440.6.1.15]
together or separately. Conf Proc IEEE Eng Med Biol Soc; Vancouver,
174. Gopura RAR, Kiguchi K: A human forearm and wrist motion assist
Canada 2008, 2008:3475–3478. [http://dx.doi.org/10.1109/IEMBS.2008.
exoskeleton robot with EMG-based Fuzzy-neuro control. In Proc.
4649954]
2nd IEEE RAS & EMBS International Conference on Biomedical Robotics and
Biomechatronics (BioRob). Scottsdale, AZ; 2008:550–555. 192. Adamovich SV, Fluet GG, Mathai A, Qiu Q, Lewis J, Merians AS: Design of
a complex virtual reality simulation to train finger motion for
175. Song R, Tong KY, Hu XL, Zheng XJ: Myoelectrically Controlled Robotic
persons with hemiparesis: a proof of concept study. J Neuroeng
System That Provide Voluntary Mechanical Help for Persons after
Rehabil 2009, 6:28. [http://dx.doi.org/10.1186/1743-0003-6-28]
Stroke. In Proc. IEEE 10th International Conference on Rehabilitation
Robotics (ICORR). Noordwijk, Netherlands; 2007:246–249. 193. Marchal-Crespo L, Reinkensmeyer DJ: Review of control strategies for
176. Li Q, Wang D, Du Z, Song Y, Sun L: sEMG Based Control for 5 DOF robotic movement training after neurologic injury. J Neuroeng
Upper Limb Rehabilitation Robot System. In Proc. IEEE International Rehabil 2009, 6:20. [http://dx.doi.org/10.1186/1743-0003-6-20]
Conference on Robotics and Biomimetics (ROBIO). Kunming, China; 194. Patton JL, Stoykov ME, Kovic M, Mussa-Ivaldi FA: Evaluation of robotic
2006:1305–1310. training forces that either enhance or reduce error in chronic
177. Mayr A, Kofler M, Saltuari L: [ARMOR: an electromechanical robot for hemiparetic stroke survivors. Exp Brain Res 2006, 168(3):368–383.
upper limb training following stroke. A prospective randomised [http://dx.doi.org/10.1007/s00221-005-0097-8]
controlled pilot study]. Handchir Mikrochir Plast Chir 2008, 40:66–73. 195. Hogan N: Impedance control: an approach to manipulation. Part 1 -
[http://dx.doi.org/10.1055/s-2007-989425] Theory. J Dynamic Syst, Meas Control 1985, 107:1–7.
178. Kawasaki H, Ito S, Ishigure Y, Nishimoto Y, Aoki T, Mouri T, Sakaeda H, 196. Ott C, Mukherjee R, Nakamura Y: Unified Impedance and Admittance
Abe M: Development of a Hand Motion Assist Robot for Control. In Proc. IEEE Int Conf Robotics and Automation (ICRA).
Rehabilitation Therapy by Patient Self-Motion Control. In Proc. IEEE Anchorage, AK; 2010:554–561.
10th International Conference on Rehabilitation Robotics (ICORR). 197. Frisoli A, Borelli L, Montagner A, Marcheschi S, Procopio C, Salsedo F,
Noordwijk, Netherlands; 2007:234–240. Bergamasco M, Carboncini MC, Tolaini M, Rossi B: Arm rehabilitation
179. Ding M, Ueda J, Ogasawara T: Pinpointed muscle force control using with a robotic exoskeleleton in Virtual Reality. In Proc. IEEE 10th
a power-assisting device: System configuration and experiment. In International Conference on Rehabilitation Robotics (ICORR). Noordwijk,
Proc. 2nd IEEE RAS & EMBS International Conference on Biomedical Robotics Netherlands; 2007:631–642.
and Biomechatronics (BioRob). Scottsdale, AZ; 2008:181–186. 198. Lo AC: Clinical designs of recent robot rehabilitation trials. Am J Phys
180. Lam P, Hebert D, Boger J, Lacheray H, Gardner D, Apkarian J, Mihailidis A: Med Rehabil 2012, 91(11 Suppl 3):S204–S216. [http://dx.doi.org/10.1097/
A haptic-robotic platform for upper-limb reaching stroke therapy: PHM.0b013e31826bcfa3]
199. Kwakkel G, Kollen BJ, Krebs HI: Effects of robot-assisted therapy on

upper limb recovery after stroke: a systematic review. Neurorehabil
Neural Repair 2008, 22(2):111–121.
200. Mehrholz J, Platz T, Kugler J, Pohl M: Electromechanical and
robot-assisted arm training for improving arm function and
activities of daily living after stroke. Cochrane Database Syst Rev 2008,
4:CD006876. [http://onlinelibrary.wiley.com/doi/10.1002/14651858.
CD006876.pub2/full]
201. Timmermans AAA, Seelen HAM, Willmann RD, Kingma H:
Technology-assisted training of arm-hand skills in stroke: concepts
on reacquisition of motor control and therapist guidelines for
rehabilitation technology design. J Neuroeng Rehabil 2009, 6:1.
[http://dx.doi.org/10.1186/1743-0003-6-1]
202. Waldner A, Tomelleri C, Hesse S: Transfer of scientific concepts to
clinical practice: recent robot-assisted training studies. Funct Neurol
2009, 24(4):173–177.
203. Lo AC, Guarino PD, Richards LG, Haselkorn JK, Wittenberg GF,
Federman DG, Ringer RJ, Wagner TH, Krebs HI, Volpe BT, Bever CTJr,
Bravata DM, Duncan PW, Corn BH, Maffucci AD, Nadeau SE, Conroy SS,
Powell JM, Huang GD, Peduzzi P: Robot-assisted therapy for
long-term upper-limb impairment after stroke. N Engl J Med 2010,
362(19):1772–1783. [http://dx.doi.org/10.1056/NEJMoa0911341]
204. Brochard S, Robertson J, Médée B, Rémy-Néris O: What’s new in new
technologies for upper extremity rehabilitation? Curr Opin Neurol
2010, 23(6):683–687. [http://dx.doi.org/10.1097/WCO.
0b013e32833f61ce]
205. Reinkensmeyer DJ, Boninger ML: Technologies and combination
therapies for enhancing movement training for people with a
disability. J Neuroeng Rehabil 2012, 9:17. [http://dx.doi.org/10.1186/
1743-0003-9-17]
doi:10.1186/1743-0003-11-3
Cite this article as: Maciejasz et al.: A survey on robotic devices for upper
limb rehabilitation. Journal of NeuroEngineering and Rehabilitation 2014 11:3.
Submit your next manuscript to BioMed Central

and take full advantage of:
• Convenient online submission

• Thorough peer review
• No space constraints or color ﬁgure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at

www.biomedcentral.com/submit

Maciejasz2014 Article ASurveyOnRoboticDevicesForUppe

Uploaded by

Copyright:

Available Formats

Maciejasz2014 Article ASurveyOnRoboticDevicesForUppe

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Maciejasz2014 Article ASurveyOnRoboticDevicesForUppe

Uploaded by

Copyright:

Available Formats

Maciejasz et al.

Journal of NeuroEngineering and Rehabilitation 2014, 11:3

R EVIEW Open Access

A survey on robotic devices for upper limb

Introduction these problems, prevailing research studies present a wide

Table 1 Robotic devices for upper limb rehabilitation

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 1 Robotic devices for upper limb rehabilitation (Continued)

Table 4 Glossary of terms concerning actuation of robots

199. Kwakkel G, Kollen BJ, Krebs HI: Effects of robot-assisted therapy on

Submit your next manuscript to BioMed Central

• Convenient online submission

Submit your manuscript at

You might also like