Background
Disease | Disease Prevalence (US cases per year) | Motor Impairment Prevalence* | Type of Upper Extremity Impairment |
---|---|---|---|
Arthritis [21] | 78 million (projected prevalence by 2040) | 3 million (2009) | Grasping |
Cerebral palsy [22] | 1 in 323 children (2008) | ~ 50% of children | Arm-hand dysfunction |
Parkinson’s Disease [23] | 500,000 (2010) | Not reported | Tremor, rigidity, akinesia/bradykinesia |
Spinal Cord Injury [24] | 282,000 (2016) | 58.3% | Tetraplegia |
795,000 (incidence, 2016) | 50% | Upper extremity hemiplegia |
Methods
Search strategy
Database | Search query |
---|---|
SportDiscus | TX “soft robot*” |
Pubmed | (“soft robot*” [All Fields] OR (“Robotics” [MeSH] AND “soft” [All Fields])) AND (hand OR finger OR thumb OR glove [All fields]) |
Scopus | TITLE-ABS-KEY(Robot*) AND TITLE-ABS-KEY(Soft) AND TITLE-ABS-KEY(“hand” OR “finger” OR “thumb” OR “glove”) |
Web of Science | TI = (soft robot*) AND TI = (hand OR finger OR thumb OR glove) |
-
The device described was a soft robot: No rigid components on the robot-human interface or minimal rigid components that will not impose physical restraints on joint motions
-
The device intended to facilitate the movement of at least one finger joint in the hand
-
The paper was a scientific article written in the English language and accessible to the authors
-
The device contained rigid components on the robot-human interface that could reasonably impose physical restraints on joint motions
-
The device focused on other joints without including the fingers
-
There was insufficient information about the device’s design such that analysis was not clear
-
The device was intended for use as a prosthetic
Framework for comparison
Soft robotic device
Environment
Mode | Description |
---|---|
Active Resistance (AR) | Patient attempts to exercise hand against a resistive force from the device |
Continuous Passive Motion (CPM) | Patient is subjected to repetitive motion by the device |
Task Specific Training (TST) | Patient is given a specific action to complete (ie grabbing a ball) while the device provides assistance |
Virtual Reality (VR) | Patient is placed in a virtual reality while the device assists in various activities |
Results: literature collection and framework summary overview
Literature collection
Overview summary of component development
Device / Group | Assisted Motion | Portability | Safety | User Intent Modality | Total DOF | No. Ind. Actuators | Weight (g) | Input force | Ext. torque / Grip force |
---|---|---|---|---|---|---|---|---|---|
Cable systems | |||||||||
Biggar et al. [27] | F | Y | – | – | 9 | 3 | – | ||
Cao et al. [28] | F | – | – | sEMG | 9 | 1 | 50 | ||
Exo-Glove Poly / Kang et al. [29] | E/F | Y | – | – | 9 | 2 | – | - / 29.5 N | |
E/F | Y | – | Bend sensors | 9 | 3 | 194 | 50 N | - / 40 N | |
GraspyGlove / Popov et al. [33] | E/F | Y | – | – | 12 | 1 | 250 | ||
GRIPIT / Kim et al. [34] | F | – | – | – | 9 | 1 | 40 | ||
F | Y | – | Pressure sensors | 9 | – | 70 | |||
E/F | Y | Spool rotational limit | sEMG | 15 | 1 | – | |||
Park et al. [40] | E/F | Y | Magnetic coupling | – | 15 | 2 | – | 34 N | - / 35 N |
RoboGlove / Diftler et al. [17] | F | Y | Multi-modal | – | 15 | 3 | 771 | - / 222 N | |
SEM Glove / Nilsson et al. [41] | F | Y | – | Pressure sensors | 9 | 3 | – | 20 N | - / 24 N |
E | Y | Verbal command | sEMG + Voice | 15 | 1 | 225 | |||
Xiloyannis et al. [46] | F | Y | – | – | 9 | 1 | – | ||
Yao et al. [47] | E/F | – | – | – | 17 | – | 85 | - / 11 N | |
Yi et al. [48] | E/F | – | – | – | 12 | – | < 100 | ||
Pneumatic systems | |||||||||
Al-Fahaam et al. [49] | F | Y | Pinky control | – | 12 | – | 100 | 400 kPa | - / 17 N |
Coffey et al. [50] | E | Y | – | EEG | 15 | 1 | – | ||
Exo-Glove PM / Yun et al. [51] | F | Y | Pressure sensor | – | 16 | 1 | – | 300 kPa | - / 22 N |
Kline et al. [19] | E | – | Pressure sensor | sEMG | 15 | 1 | 100 | 34 kPa | < 1 Nm / - |
Li et al. [52] | E | – | – | – | 15 | 1 | – | ||
Low et al. [53] | F | – | – | – | 3 | 1 | 25 | ||
Maeder-York et al. [54] | F | Y | – | – | 3 | 1 | – | 207 kPa | |
F | Y | – | – | 12 | – | 180 | 120 kPa | - / 41 N | |
Nordin et al. [56] | F | – | Emergency button | – | 15 | 3 | – | 200 kPa | - / 3.61 N |
Noritsugu et al. [57] | F | – | – | – | 15 | 2 | 120 | 500 kPa | |
PneuGlove / Connelly et al. [20] | E | – | Bend sensor | – | 15 | 5 | 68.9 kPa | 2.7 Nm / - | |
Polygerinos et al. [58] | F | – | – | – | 12 | 1 | 160 | 43 kPa | - / 4.42 N |
Power Assist Glove / Toya et al. [18] | F | – | – | Bend sensors | 15 | 4 | 180 | ||
PowerAssist Glove / Kadowaki et al. [59] | E/F | – | – | sEMG | 15 | – | 135 | ||
RARD / Chua et al. [60] | Abduction/Adduction | – | – | – | 1 | 1 | – | ||
F | – | Pressure sensor | – | 15 | 5 | – | 50 kPa | ||
Reymundo et al. [65] | E | – | – | – | 3 | 1 | – | 50 kPa | |
Tarvainen et al. [66] | F | – | – | – | 3 | 2 | – | ||
Wang et al. [67] | E/F | – | – | – | 15 | 5 | – | 675 kPa | - / 21.24 N |
Wang et al. [68] | F | – | – | – | 3 | 1 | – | 350 kPa | |
F | Y | – | sEMG | 15 | 5 | 170 | 120 kPa | - / 6.5 N | |
F | – | – | – | 12 | 1 | 200 | |||
E/F | Y | – | – | 15 | 5 | 180 | 120 kPa | - / 8.4 N | |
Yap et al. [74] | E | Y | – | – | 15 | 5 | 150 | 100 kPa | 4.25 Nm / - |
Yap et al. [75] | F | – | – | – | 3 | 1 | – | 200 kPa | |
Yeo et al. [76] | F | – | Strain sensor | – | 3 | 1 | – | 110 kPa | |
Zaid et al. [77] | F | – | – | – | 6 | 2 | – | ||
Zhang et al. [78] | F | – | – | – | 3 | 1 | – | ||
Hydraulic systems | |||||||||
F | Y | Emergency button | sEMG | 15 | 5 | 285 | 413 kPa | - / 14.15 N |
Discussion of framework elements
Soft robotic device design
Portability
Safety input
Safety Mechanism | Description | Advantages | Disadvantages |
---|---|---|---|
Spool rotational limit | Spools which guide the cables are limited in rotation, thereby preventing hyper-flexion/ extension | Intrinsically built into the system to avoid over-actuation of the digits. | Possibility of failure if patient initiates device mode incorrectly |
Pressure sensor | Pressure sensor shuts off actuation when threshold pressure is exceeded | Sensor can be easily incorporated into control unit | Pressure thresholds may not be the same among patients with differing degrees of impairment |
Emergency button | A button is available on the control unit to provide immediate cessation of actuation | Patient has ability to override the device when they sense discomfort | Patients with impairments may not react quickly to prevent damage from severe malfunction |
Bend sensors / Strain sensors | Sensors placed along the joints can detect and control the degree of bending | Can more directly measure the degree of joint bending | May be more difficult to implement and must be cautious when adding them to the hand orthosis |
Unactuated digit detection | Monitors the movement of a digit that is not actuated so that the patient’s voluntary movement of that digit sends a signal to the device to turn off | Patient has ability to determine when to shut off the device | Requires residual function in a digit, forces device to leave at least one digit unactuated |
Magnetic coupling | The actuator cables are magnetically coupled to the robotic tendons and detach when the tension is too high | Patient does not need to alert for termination | May be difficult to customize for varying levels of hand dysfunction |
Verbal command | The user says a verbal command, such as “stop” | Patient can quickly terminate device | Voice recognition failure |
Multi-modal feedback | Sensors for temperature, motor current, battery levels, and loss of sensor feedback all have the ability to cease operation | Many layers of security to greater ensure protection from electrical components of device | Only motor current to the actuators is vaguely correlated to degree of finger movement |
Feedback input
Feedback Modality | Description | Advantages | Disadvantages |
---|---|---|---|
Bend Sensors – Digits | Sensors are placed on all finger joints and a joint pattern analysis can detect a user’s specific intended hand motion | Is able to differentiate specific hand motions and does not require electrode placement by the patient | Cannot be used in patients with complete hand paralysis |
Pressure Sensors - Digits | |||
Bend Sensors – Wrist | A bend sensor is placed on the wrist as wrist motion is likely still a familiar motion in patients with hand impairment | Simple to implement and can reliably detect wrist motion. Does not require electrode placement by the patient | May not be able to distinguish specific hand motions and requires wrist motion to be intact |
EEG | An EEG pattern analysis was obtained on healthy patients in order to be able to identify similar patterns in patients with hand paralysis | Can be implemented in a patient with complete paralysis because acquires signal for intent at the beginning of motor pathway | Requires many electrodes to be placed on the head and may be the least reliable means of detection of user intent of those presented |
sEMG | Electrodes are placed on major muscles of the forearm to detect myoelectric activity in order to gauge user intent | Reliably detects forearm activity and is able to differentiate some specific hand motions | Requires some residual level of muscle activity |
Voice activated | Voice commands can operate the device | Unambiguously controls the device | Not a part of neuromuscular pathway so effects on neuroplasticity are less clear |
Human-robot interface
Interaction with environment
Methods of evaluation
Measurement | Description |
---|---|
Extension torque | The torque applied by the device on finger extension |
Grasping ability | Tested whether subject was able to grab various objects with assistance from the device |
Grip force | The force exerted by the device attempting a grasping motion with subject completely passive |
Max input force | Either the max input force supported by the device or the max input force required to achieve the desired functionality (pneumatic and hydraulic systems only) |
Motion trajectory | Tracks the trajectory of the device/digits upon actuation |
Opposition grasp force | The actuated force achieved while opposing the thumb |
Pinch force | The force exerted by the device attempting a pinching motion with subject completely passive |
ROM | Measurement of the rotations about the joints in the hands |
Speed of movement | Speed of movement of the fingertip upon actuation |
Tensile force | The max tension required to achieve desired function (cable systems only). It is the equivalent to max input force of pneumatic systems. |