0
Review Article

Development of a Dexterous Prosthetic Hand OPEN ACCESS

[+] Author and Article Information
Mahonri Owen

Faculty of Engineering,
University of Waikato,
Hamilton 3216, New Zealand
e-mail: mahonri.owen@gmail.com

ChiKit Au

Faculty of Engineering,
University of Waikato,
Hamilton 3216, New Zealand
e-mail: ckau@waikato.ac.nz

Andrew Fowke

Faculty of Engineering,
University of Waikato,
Hamilton 3216, New Zealand
e-mail: andrewfowke@hotmail.com

1Corresponding author.

Contributed by the Computers and Information Division of ASME for publication in the JOURNAL OF COMPUTING AND INFORMATION SCIENCE IN ENGINEERING. Manuscript received January 28, 2016; final manuscript received October 11, 2017; published online November 13, 2017. Assoc. Editor: Charlie C. L. Wang.

J. Comput. Inf. Sci. Eng 18(1), 010801 (Nov 13, 2017) (7 pages) Paper No: JCISE-16-1045; doi: 10.1115/1.4038291 History: Received January 28, 2016; Revised October 11, 2017

An anthropomorphic, under-actuated, prosthetic hand has been designed and developed for upper extremity amputees. This paper proposes a dexterity focused approach to the design of an anthropomorphic electromechanical hand for transradial amputees. Dexterity is increased by the improvement of thumb position, orientation, and work space. The fingers of the hand are also capable of adduction and abduction. It is the intent of this research project to aid the rehabilitation of upper extremity amputees by increasing the amount of tasks the hand can execute. Function and control of the hand are based on micro servo actuation and information acquired from the brain. Electroencephalography (EEG) is used to attain the mental state of the user, which triggers the prosthetic hand. This paper focuses on the mechanical arrangement of the hand and investigates the effect of increasing the degrees-of-freedom (DOFs) the thumb and fingers have.

FIGURES IN THIS ARTICLE
<>

Estimates suggest that ten million people on the earth at any one point in time suffer from the effects of a missing limb or body part [1]. Thirty percent of these people are upper extremity amputees that suffer from the loss of either their whole arm or parts of it. Until recent years, the development of prosthetic devices that return function and confidence to these amputees has been very limited. Over the last decade, new information about prosthetic development and design has been brought to light. Never before have we been able to mimic the aesthetics, function, and performance of a human hand as we can today. In New Zealand, there is a serious gap in the knowledge required for the design and control of anthropomorphic robotic prosthetic hands. This knowledge will only be gained by the research, design, and development of these devices in the south pacific.

The aim of the project is to develop an anthropomorphic, prosthetic hand which can perform the basic functions of a human natural hand. The mechanical design and electronic control of an artificial anthropomorphic hand require interdisciplinary research in the fields of electrical engineering, mechanical engineering, computer science, economics, and mathematics. An integrated design approach between mechanics and electronic control applied to an under-actuated artificial hand for prosthetic applications will be presented. The major contribution of this research is to support the rehabilitation of amputees. One of the issues for an amputee using a prosthetic limb is the limited amount of tasks the hand is able to perform and execute. This research will increase the availability of prosthetic devices and encourage the development of functional, dexterous, and useful prosthetic devices for upper extremity amputees.

The human hand is made up of twenty-seven bones, which can be described in three layers. The first layer of bones is the carpals, the second layer of bones is the metacarpals, and the third layer of bones is the phalanges. There are usually eight carpal bones found in the wrist, which are connected to the metacarpal bones by carpometacarpal (CMC) joints. There are five metacarpal bones located in the palm, which are connected to the phalange bones by metacarpophalangeal joints.The digits of the hand consists of fourteen phalange bones: three per finger and two for the thumb. Each finger has a proximal phalange, an intermediate phalange, and a distal phalange. The proximal phalange is connected to the intermediate phalange by the proximal interphalangeal joint. The intermediate phalange is connected to the distal phalange by the distal interphalangeal (DIP) joint. The thumb has a proximal phalange and a distal phalange which are connected by the interphalangeal joint.

The amputation of a hand causes severe disruption to an amputee's life, both physically and physiologically. In order for the amputee to return to living a healthy lifestyle, it is essential to have a prosthetic to aid in simple tasks. The prosthetic should return as much function and appearance as possible [2]. Many prosthetic hands such as iLimb, Bebionic, and Michelangelo are very similar in design to the human hand with five digits. Each also has a silicon glove to help with anthropomorphism and grip.

The degree-of-freedom (DOF) in the human hand is large, as many as twenty-one can be counted [3]. Based on the fact that joints in the human hand work together to perform and execute grips, the kinematic structure of the prosthetic hand can be simplified. One obvious simplification is the coupling of the finger joints to allow flexion and extension. The complexity of the thumb makes it very hard to replicate all DOF in limited spaces. It is concluded that two independent degrees-of-freedom would be adequate to achieve multiple grasping tasks, replicating flexion/extension and adduction/abduction [4].

Independent, serial, and parallel joints are three common types of joints used in prosthetic hands. They differ from each other by the way they are connected to an actuator. An independent joint has one dedicated actuator assigned to it, coupled either directly or indirectly.

An independent joint has a transmission device between actuator and joint which plays key roles in reducing output speed and increasing torque. To some degree, it also absorbs impact force. A serial-driven joint is also called a coupling joint, which has several joints driven by one actuator. Movements of the chain of joints connected to one actuator are all associated with each other where only the proximal joint is coupled with the actuator and all following joints are coupled with the proximal joint using transmission structures such as link bars and pulleys. Joint coupling can be realized by using tendons where all joints within the finger are connected by pulley and belt. These joints bend with the same velocity. It can also be implemented by using linkage bars with a fixed DIP joints. Using linkage bars, a very stiff structure is created, which is good at precision movements. The fingers on many commercial products in the past are a fixed shape with the only joint being the metacarpophalangeal replicate; these are independent joints controlled by one actuator each. The kinematic structure of these designs is simplified; however, the contact area between object and hand is small and results in an inadequate grip. Parallel-driven joints are those where several joints are independent in logic and driven by only one dedicated actuator. This is also called the single-motor-driven concept [5] and has been applied in prosthetic hands in the past where all fingers move at the same time (usually with 1 or 2 finger grippers only).

The control of modern prosthetics varies and is wide ranging. There are multiple methods employed to control prosthetic hands; one common way is the electric switch in combination with cables or straps. This method requires training and is used together with body movement to create movement in the prosthesis. For instance, a switch is pushed by protraction to extend the arm, and retraction of the shoulder would hit another switch to flex the arm while elevation opens up the fingers of a prosthetic hand to perform a task. These control methods can be cumbersome to operate and are limited in grasping ability. More importantly, these simple motions do not mimic normal fluid human motion.

Another approach used for controlling prosthetic hands is using electromyography (EMG) signals. EMG signals are acquired from muscles excited from the peripheral nervous system. This approach does not require extra movement and are more precise than electric switches. Most researches on EMG signal-controlled prosthetic hand rely on either the EMG pattern recognition or the information of force level of the limb motion [6]. Various methods such as neural networks [713] and wavelet transform [14,15] are proposed to extract and recognize the features of EMG patterns. However, difficulties arise due to variations in patterns from user to user and differences in electrode positions. Furthermore, the time variation may not be consistent due to the fatigue and sweat of the arm. These cause difficulties in controlling the prosthetic hand to perform accurate and dexterous movements.

As a result, other controlling approaches such as voice-controlled prosthetic hands [16] are proposed. Although this method is simple to implement and no additional training is required to use the prosthetic, giving a voice signal before performing a grasp is unusual and out of the normal. In fact, many users of these prosthetic devices feel like they do not fit into society due to the unnatural feeling of voice controlling their prosthesis.

Electroencephalography (EEG) signals can also be used for controlling and actuating prosthetics. The brain signals are generally categorized into frequency bands and are given names: Some commonly accepted names are Alpha, Beta, Gamma, Delta, and Theta. The Alpha frequencies are further split into Alpha 1 and Alpha 2, which have the values 8–9 Hz and 10–12 Hz, respectively. EEG is the recording of electrical activity along the scalp. The process involves monitoring cells in the brain called neurons for action potential. An action potential is an electrical impulse produced from the neuron and is what neurons use to communicate with each other. The electrical impulses represent the information carried over millions of neurons. Recording areas of neuronal interaction produces signals that can be compartmentalized and used to represent specific desired actions of the person who is subject to the recording of brain signals. Commands for actuating the prosthetics are issued to the prosthetics after analyzing the brain signal to determine the desired action.

The natural complexity of the human hand necessitates simplification when attempting to mechanically replicate its function and dexterity. Actuator size and hand geometry determine the degree of simplification required. Mechanical simplicity facilitates the control of the hand but has an adverse and unwanted effect on the dexterity; therefore, a relationship between complexity and function exists. The balance between complexity and function is of utmost importance when designing a prosthetic hand.

Due to the limited space within the hand, a self-contained hand with its actuators located within the hand usually possesses low degree-of-freedom. Table 1 lists the degree-of-freedom of current prosthetic hands and whether they are self-contained or not.

The prosthetic hand consists of three categories of components: the fingers which include the index, middle, ring and little fingers, the thumb and the palm. Figure 1 shows a finger including all finger bones using servo motor as actuator. Two link bars are highlighted so that both DIP and proximal interphalangeal are coupling joints. The servo motor is installed at the base of the finger. A nylon stri ng is used to connect the motor with the proximal phalanges.

The thumb is a very important aspect of the hand design as two motors are required to control it. Figure 2 shows how the two motors are positioned and employ gears to transmit the power from the servo.

Servo 1 as shown, which is held within the palm of the hand, is attached to a gear. The range of motion for this joint is about 130˚ from the horizontal so that the thumb CMC joint would have to be very sleek in order to not interfere with servo 1. Housing servo 2 within the proximal bone of the thumb implies that the mechanism for controlling the flexion of the intermediate bone could be simplified enormously. There is coupling within the thumb which is similar to the fingers; however, only one link is required.

Figure 3 shows the palm which holds two servos. One is for the middle finger and the other is for controlling the adduction/abduction of the fingers. It also has three holes to accommodate servo holders of the index finger, ring finger, and little finger. A special bay is at the lower right-hand side to hold the thumb. Figure 4 shows the adduction and abduction movement of the fingers after installation on the palm. Figure 4(a) shows how the fingers are positioned when they are not spread. A nylon string is attached to each finger except the middle finger which is stationary. A hole is cut into the palm which the nylon strings from each finger are fed to the back of the hand where the servo is positioned. Figure 4(b) shows the hand with the fingers in the abduction position.

The function of a prosthetic hand refers to how close it can perform a grasping task in relation to the human hand. This depends on the number of joints, their positions, and their orientations.

One of the major factors affecting the function of a prosthetic hand is the orientation of the thumb. Many human hand tasks cannot be performed without an appropriate thumb orientation. A matlab toolbox for comparing the functionality of various robotic and prosthetic hands with a human hand was developed by Feix [17]. An action manifold space that all the postures of a human hand can reach is determined by placing sensors on each fingertip and the hand dorsum. Similarly, the action manifold of a prosthetic hand is obtained by using the forward kinematics equations. Each phalange is connected by a revolute joint based at the joint of the previous phalange. Each joint is coupled to the previous joint; therefore, movement of the proximal phalange induces movement on the intermediate and distal phalanges. An anthropomorphism index [18], which is the overlap of the action manifold of the prosthetic hand and human hand, is computed using the matlab toolbox. This index can be used as a measure of how similar a prosthetic hand can function as a human hand.

The thumb orientation is defined by the angle between the thumb face normal and the projection on the major thumb axis as shown in Fig. 5(a). A set of anthropomorphism indices of the designed prosthetic hands with various thumb orientations is computed. Figure 5(b) plots these anthropomorphism indices against the thumb orientation.

It is found that the thumb orientation of about 40 deg yield the maximum anthropomorphism index, which implies the highest similarity in terms of functionality between the prosthetic and human hand. Figure 6(a) is the Denavit–Hartenberg representation plotted by the toolbox [17] while Fig. 6(b) shows the computer-aided design model of the prosthetic hand.

The prosthetic hand components are printed three dimensionally. The link bars are manufactured by laser cutting. The components are assembled together and a microcontroller is attached to the prosthetic hand. The FSR are installed at the finger tip to provide feedback on the status of the grasp. Two FSR are positioned on the thumb as shown in Fig. 7. One is located on the thumb distal and the other on the thumbs proximal phalange.

The microcontroller controls the servos to replicate a human grasp. For instance, Fig. 8 depicts the steps for the prosthetic hand to grasp a ball using three fingers. There are seven servos in the prosthetic hand as numbered in the figure. Initially, servo 7, which is located on the palm, is actuated so that the prosthetic fingers adducts. Then servos 5 and 6, which control the flexion of the little and ring finger, are commanded to close fully. Finally, servos 3 and 4 for index and middle finger and servos 1 and 2 for the thumb are actuated to close fully so that the ball is grasped securely.

Various servo actuation combinations and orders are defined for different grasping tasks. Figure 9 shows another three-finger grasp with a different sequence of servo motor actuations.

The dexterity of a prosthetic hand is very important; amputees need a prosthetic to provide a wide range of options to grip both small and large objects. The designed hand has the ability to have the thumb to interact with all the other digits. This is one sure way of proving high prosthetic hand dexterity. The fingers can abduct and adduct, which also increase the dexterity of the hand. Figure 10 shows the fingertip and thumb trajectories of the prosthetic hand.

The thumb is capable of interacting with each finger of the hand by the intersecting trajectories of the thumb (black) with the fingers (remaining colors). Maximizing the grip strength is possible through thumb opposition to each finger. The large workspace of the thumb increases the amount of objects that can be gripped. Figure 11 physically shows this ability without any interference.

The hand is capable of performing finger flexion and extension, finger abduction and adduction, thumb flexion and extension, thumb abduction and adduction, and thumb radial abduction and adduction. This is not seen in any other prosthetic hands in the market to date. This ability increases the three-dimensional workspace of the thumb immensely. The thumb is able to navigate across the three Cartesian coordinates and interact closely with the fingers.

Although the mechanical design of the hand has been improved and shows anthropomorphism, there are still limitations to its operation and function. The thumb as currently designed is not capable of performing lateral grasps. The lateral grasp is used in everyday life and is used to hold things like key and credit cards. Improvement can be made by introducing a third degrees-of-freedom to the thumb allowing complex multi-axial movement at the CMC joint. This change would necessitate a complete rework of the actuation system. With the current design, it is not feasible to design a three degrees-of-freedom thumb with the unavoidable mechanical off set in the thumb joint created by the servo motors. The thumb needs to be supported by a saddle joint and actuated by servo motors that are not contained within the hand. This design implies the use of tendons to actuate the thumb.

The prosthetic hand is controlled by the brain. It is currently using the Mindwave mobile headset by Neurosky to monitor the electroencephalography signals emitted from neuronal activity in the brain and correlates the data into useable information. The signals acquired by the headset are deciphered using proprietary algorithms designed by Neurosky. The headset is able to extract the features of the signals and deduce the level of concentration of an individual. Concentration is measured on a scale from zero to one hundred, where one hundred represents complete concentration and zero represents no concentration. The output of the headset is then translated into servo actuation commands for controlling the prosthetic hand.

The outgoing data from the headset will be used to trigger a command on the microboard. The command will cause the hand to perform the ball grasp. For instance, an elevated level of concentration higher than eighty-five will close the hand while an attention state lower than eighty-five percent will open the hand. Once the grasp has been completed, the hand will return to a relaxed position awaiting the signal to close or remain open. Wireless communication to the prosthetic hand is accomplished through the use of the DF-BluetoothV3 Bluetooth module.

Figure 12 explains the current implementation of prosthetic control starting with user input and ending in prosthetic response. The process consists of four steps. First, EEG acquisition occurs through two electrodes. One electrode is placed on the ear and the other is placed on the forehead above the right eye. Both electrodes are dry and require direct contact with the skin. One electrode acts as the base signal with the other providing EEG information. Second, signal analysis occurs in the headset. The headset is capable of identifying two brain states: concentration and meditation. The headset is also able to detect intentional blinking. For the purpose of this paper only the concentration of the user is monitored in real time. Third, data communication occurs through direct communication from the headset to the Arduino microcontroller. The Arduino receives real-time data pertaining to the users' mental state. Real-time data are received as an integer value that determines servo position. Finally, prosthetic actuation occurs through servo motors. When a signal is identified by the Arduino with elevated concentration, a command is given to the servo motors to activate. Servo actuation closes the fingers of the prosthetic hand. Force sensors on the fingers of the prosthetic hand provide feedback to the Arduino, determining the force at which the hand is to grasp the ball.

An anthropomorphic, under-actuated, prosthetic hand has been designed and developed for upper extremity amputees. Function of the hand is based on micro servo actuation. This paper focuses on the mechanical arrangement of the hand and evaluates its effectiveness in grasping a ball. This is important because it relates directly to the aim of the article to create a dexterous prosthetic hand.

The hand is responsive to signals emitted from the brain and is actuated by the mental state of the user. Improvement can come from refining the control of the system to respond to an intention to act by the amputee. Further mechanical improvements include implementing a biologically inspired design based on the human hands bones, actuating the hand through tendons, and increasing the range of motion of the thumb by introducing a saddle joint that will allow complex multi-axial movement.

LeBlanc, M. , 2011, “ Give Hope—Give a Hand,” The Ellen Meadow Prosthetic Hand Foundation, San Francisco, CA, accessed June 5, 2017, http://web.stanford.edu/class/engr110/2011/LeBlanc-03a.pdf
Kady, A. , Mahfouz, A. , and Taher, M. , 2010, “ Mechanical Design of an Anthropomorphic Prosthetic Hand for Shape Memory Alloy Actuation,” Fifth International Conference on Biomedical Engineering (CIBEC), Cairo, Egypt, Dec. 16–18, pp. 86–89.
Li, S. , Sheng, X. , Zhang, J. , and Xiangyang, Z. , 2011, “ Design of an Anthropomorphic Prosthetic Hand Towards Neural Interface Control,” Fifth International Conference on Intelligent Robotics and Applications (ICIRA), Aachen, Germany, Dec. 6–8, pp. 507–517.
Light, C. , and Chappell, P. , 2000, “ Development of a Lightweight and Adaptable Multiple-Axis Hand Prosthesis,” Med. Eng. Phys., 22(10), pp. 679–684. [CrossRef] [PubMed]
Liu, Y. , Li, S. , and Xie, M. , 2007, “ Design and Implementation of a New Single-Motor Driven Arm Manipulator,” IEEE International Conference on Mechantronics and Automation (ICMA), Harbin, China, Aug. 5–8, pp. 3071–3076.
Zhao, J. , Xie, Z. , Jiang, L. , Cai, H. , Liu, H. , and Hirzinger, G. , 2006, “ A Five-Fingered Underactuated Prosthetic Hand Control Scheme,” The First IEEE/RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob), Pisa, Italy, Feb. 20–22, pp. 995–1000.
Kelly, M. F. , Parker, P. A. , and Scott, R. N. , 1990, “ The Application of Neural Networks to Myoelectric Signal Analysis: A Preliminary Study,” IEEE Trans. Biomed. Eng., 37(3), pp. 221–230. [CrossRef] [PubMed]
Hiraiwa, A. , Shimohara, K. , and Tokunaga, Y. , 1989, “ EMG Pattern Analysis and Classification by Neural Network,” IEEE International Conference on Systems, Man and Cybernetics, Cambridge, MA, Nov. 14–17, pp. 1113–1115.
Koike, Y. , and Kawato, M. , 1994, “ Estimation of Arm Posture in 3D-Space From Surface EMG Signals Using a Neural Network Model,” Trans. Inst. Electron., Inf., Commun. Eng., J77-D-II(1), pp. 193–203.
Farry, K. A. , Walker, I. D. , and Baraniuk, R. G. , 1996, “ Myoelectric Teleoperation of a Complex Robotic Hand,” IEEE Trans. Rob. Autom., 12(5), pp. 775–787. [CrossRef]
Huang, H.-P. , and Chen, C.-Y. , 1999, “ Development of a Myoelectric Discrimination System for a Multi-Degree Prosthetic Hand,” IEEE International Conference Robotics and Automation (ICRA), Detroit, MI, May 10–15, pp. 2392–2397.
Tsuji, T. , Ichinobe, H. , Ito, K. , and Nagamachi, M. , 1993, “ Recognition of Forearm Motions From EMG Signals by Error Back Propagation Typed Neural Network Using Entropy,” Trans. Soc. Instrum. Control Eng., 29(10), pp. 1213–1220. [CrossRef]
Fukuda, O. , Tsuji, T. , and Kaneko, M. , 1997, “ Pattern Classification of EMG Signals Using Neural Networks During a Series of Motions,” Trans. Inst. Elect. Eng. Jpn., 117(10), pp. 1490–1497.
Zhang, X. , Yang, Y. , Xu, X. , and Zhang, M. , 2002, “ Wavelet Based Neuro-Fuzzy Classification for EMG Control,” IEEE International Conference on Communications, Circuits and Systems and West Sino Expositions, Chengdu, China, June 29–July 1, pp. 1087–1089.
Cai, L. , Wang, Z. , and Zhang, H. , 1999, “ An EMG Classification Method Based on Wavelet Transform,” First Joint Engineering in Medicine and Biology, 21st Annual Conference and the 1999 Annual Fall Meeting of the Biomedical Engineering Society Conference (BMES/EMBS), Atlanta, GA, Oct. 13–16, p. 565.
Asyali, M. H. , Yilmaz, M. , Tokmakci, M. , and Mittal, R. , 2011, “ Design and Implementation of a Voice-Controlled Prosthetic Hand,” Turk. J. Electr. Eng. Comput. Sci., 19(1), pp. 33–46.
Feix, T., 2012, “Grade Your Hand Toolbox,” Feix, T., New Haven, CT, accessed Feb. 8, 2016, http://grasp.xief.net/toolbox.htm
Feix, T. , Romero, J. , Ek, C. H., Schmiedmayer, H.-B., and Kragic, D. , 2013, “ A Metric for Comparing the Anthropomorphic Motion Capability of Artificial Hands,” IEEE Trans. Rob., 29(1), pp. 82–93. [CrossRef]
Copyright © 2018 by ASME
View article in PDF format.

References

LeBlanc, M. , 2011, “ Give Hope—Give a Hand,” The Ellen Meadow Prosthetic Hand Foundation, San Francisco, CA, accessed June 5, 2017, http://web.stanford.edu/class/engr110/2011/LeBlanc-03a.pdf
Kady, A. , Mahfouz, A. , and Taher, M. , 2010, “ Mechanical Design of an Anthropomorphic Prosthetic Hand for Shape Memory Alloy Actuation,” Fifth International Conference on Biomedical Engineering (CIBEC), Cairo, Egypt, Dec. 16–18, pp. 86–89.
Li, S. , Sheng, X. , Zhang, J. , and Xiangyang, Z. , 2011, “ Design of an Anthropomorphic Prosthetic Hand Towards Neural Interface Control,” Fifth International Conference on Intelligent Robotics and Applications (ICIRA), Aachen, Germany, Dec. 6–8, pp. 507–517.
Light, C. , and Chappell, P. , 2000, “ Development of a Lightweight and Adaptable Multiple-Axis Hand Prosthesis,” Med. Eng. Phys., 22(10), pp. 679–684. [CrossRef] [PubMed]
Liu, Y. , Li, S. , and Xie, M. , 2007, “ Design and Implementation of a New Single-Motor Driven Arm Manipulator,” IEEE International Conference on Mechantronics and Automation (ICMA), Harbin, China, Aug. 5–8, pp. 3071–3076.
Zhao, J. , Xie, Z. , Jiang, L. , Cai, H. , Liu, H. , and Hirzinger, G. , 2006, “ A Five-Fingered Underactuated Prosthetic Hand Control Scheme,” The First IEEE/RAS-EMBS International Conference on Biomedical Robotics and Biomechatronics (BioRob), Pisa, Italy, Feb. 20–22, pp. 995–1000.
Kelly, M. F. , Parker, P. A. , and Scott, R. N. , 1990, “ The Application of Neural Networks to Myoelectric Signal Analysis: A Preliminary Study,” IEEE Trans. Biomed. Eng., 37(3), pp. 221–230. [CrossRef] [PubMed]
Hiraiwa, A. , Shimohara, K. , and Tokunaga, Y. , 1989, “ EMG Pattern Analysis and Classification by Neural Network,” IEEE International Conference on Systems, Man and Cybernetics, Cambridge, MA, Nov. 14–17, pp. 1113–1115.
Koike, Y. , and Kawato, M. , 1994, “ Estimation of Arm Posture in 3D-Space From Surface EMG Signals Using a Neural Network Model,” Trans. Inst. Electron., Inf., Commun. Eng., J77-D-II(1), pp. 193–203.
Farry, K. A. , Walker, I. D. , and Baraniuk, R. G. , 1996, “ Myoelectric Teleoperation of a Complex Robotic Hand,” IEEE Trans. Rob. Autom., 12(5), pp. 775–787. [CrossRef]
Huang, H.-P. , and Chen, C.-Y. , 1999, “ Development of a Myoelectric Discrimination System for a Multi-Degree Prosthetic Hand,” IEEE International Conference Robotics and Automation (ICRA), Detroit, MI, May 10–15, pp. 2392–2397.
Tsuji, T. , Ichinobe, H. , Ito, K. , and Nagamachi, M. , 1993, “ Recognition of Forearm Motions From EMG Signals by Error Back Propagation Typed Neural Network Using Entropy,” Trans. Soc. Instrum. Control Eng., 29(10), pp. 1213–1220. [CrossRef]
Fukuda, O. , Tsuji, T. , and Kaneko, M. , 1997, “ Pattern Classification of EMG Signals Using Neural Networks During a Series of Motions,” Trans. Inst. Elect. Eng. Jpn., 117(10), pp. 1490–1497.
Zhang, X. , Yang, Y. , Xu, X. , and Zhang, M. , 2002, “ Wavelet Based Neuro-Fuzzy Classification for EMG Control,” IEEE International Conference on Communications, Circuits and Systems and West Sino Expositions, Chengdu, China, June 29–July 1, pp. 1087–1089.
Cai, L. , Wang, Z. , and Zhang, H. , 1999, “ An EMG Classification Method Based on Wavelet Transform,” First Joint Engineering in Medicine and Biology, 21st Annual Conference and the 1999 Annual Fall Meeting of the Biomedical Engineering Society Conference (BMES/EMBS), Atlanta, GA, Oct. 13–16, p. 565.
Asyali, M. H. , Yilmaz, M. , Tokmakci, M. , and Mittal, R. , 2011, “ Design and Implementation of a Voice-Controlled Prosthetic Hand,” Turk. J. Electr. Eng. Comput. Sci., 19(1), pp. 33–46.
Feix, T., 2012, “Grade Your Hand Toolbox,” Feix, T., New Haven, CT, accessed Feb. 8, 2016, http://grasp.xief.net/toolbox.htm
Feix, T. , Romero, J. , Ek, C. H., Schmiedmayer, H.-B., and Kragic, D. , 2013, “ A Metric for Comparing the Anthropomorphic Motion Capability of Artificial Hands,” IEEE Trans. Rob., 29(1), pp. 82–93. [CrossRef]

Figures

Grahic Jump Location
Fig. 1

The finger of a prosthetic hand

Grahic Jump Location
Fig. 4

Adduction/abduction position: (a) adduction and (b) abduction

Grahic Jump Location
Fig. 5

The anthropomorphism index against the thumb orientation: (a) thumb orientation and (b) anthropomorphism index versus thumb orientation

Grahic Jump Location
Fig. 6

Prosthetic hand with thumb orientation of 40 deg: (a) Denavit–Hartenberg representation and (b) computer-aided design model

Grahic Jump Location
Fig. 7

Force sensing resistors (FSR) sensors on the thumb

Grahic Jump Location
Fig. 8

Steps for a three-finger ball grasp

Grahic Jump Location
Fig. 9

Three-finger ball grasp in fingertips

Grahic Jump Location
Fig. 10

Trajectory plot of various fingers

Grahic Jump Location
Fig. 11

The thumb touching various fingers: (a) thumb and little finger, (b) thumb and ring finger, (c) thumb and middle finger, and (d) thumb and index finger

Grahic Jump Location
Fig. 12

Control flow for the designed prosthetic hand

Tables

Table Grahic Jump Location
Table 1 The DOF of various prosthetic hands available on the market

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In