O&P Library > Atlas of Limb Prosthetics > Chapter 6C

Chapter 6C - Atlas of Limb Prosthetics: Surgical, Prosthetic, and Rehabilitation Principles

Upper-Limb Prosthetics: Components for Adult Externally Powered Systems

Craig W. Heckathorne, M.S.E.E. 

During the last decade, externally powered components have been used with increasing frequency in upper-limb fittings. There is also evidence that the percentage of persons continuing to use prostheses having these components has increased. Many factors have been suggested as contributing to the increase in clinical utilization of electric-powered components, but four factors stand out as contributing to the increase in the numbers of successful implementations, as measured by continued use of a prosthesis:

• Technological advances in actuators, materials, and controllers
• Conceptual advances leading to designs with improved performance characteristics
• The accumulation of a body of experience guiding successive clinical fittings
• The willingness of a diverse community of prosthe-tists, engineers, therapists, designers, physicians, social workers, and exemplary users to share their knowledge and experience

Community knowledge is fundamental. It is the pool into which individual accomplishments flow and from which the art and science of prosthetics, as a field of endeavor, is nourished. This section on electric-powered components is drawn from that pool of knowledge and thus represents the contributions of many individuals. It is, of necessity, a distillation intended to acquaint the reader with the topic. The content has been selected to emphasize design aspects influencing the performance and use of these components and is limited to components intended for adults, commercially produced, and readily available in North America.

The text is divided into four sections: prehension mechanisms, wrist mechanisms, enhancements to body-powered elbows, and elbow mechanisms. (There are no commercially available electric-powered shoulder components.) Each section begins with a general description of the components to be covered. Common characteristics and features are described, and where data are available, comparisons are drawn to the physiologic counterpart for which the device is intended as a replacement. Following the general description are detailed subsections describing each of the components available within the category. The component sections include construction and mechanical specifications for each device, performance characteristics, and control systems offered by the manufacturer of the device, as well as compatible control systems offered by other manufacturers.

Writing on this topic is hampered by the absence of a standardized terminology. The difficulty is most evident in the variety of descriptive names given to the control systems, even among those that are essentially similar in character. Further confusion results when the same word is used in different contexts, such as the term "proportional." One manufacturer uses the term in the context of "time proportional," to indicate that the response of the device to the control signal is proportional to the duration of time the signal is applied. Other manufacturers use the same term to denote that the action of the device is proportionally determined by the amplitude of the control signal. Both uses are technically accurate, but the controllers differ significantly in the capability they offer to regulate the action of the device.

An effort has been made throughout the discussions of control options to clarify differences and similarities. In addition, a common control terminology is used in association with the name assigned by the manufacturer to identify the number of distinct control sources and distinct device functions. For example, the Otto Bock "digital two site" myoelectric controller is also described in the text as a two-site, two-function controller to indicate that two separate and independent muscle sites are required to operate the controller and that two functions (e.g., "open" and "close") can be controlled voluntarily. The "off" condition is generally assumed, unless noted otherwise, because it is not practical to have a battery-powered device continuously "on" in the absence of a control signal. In the case of control by means other than muscle signals or in cases where myoelectric control is one of several options, the more general term "source" is used in place of "site."

In keeping with the intention of this section as a component review, techniques for incorporating components into prosthetic systems and for fabrication of prostheses are not covered. The reader is referred to the technical manuals and courses offered by the various manufacturers. Techniques for integrating multiple systems into a single prosthesis and for designing hybrid systems combining body-powered and electric-powered componentry are also not discussed. These are areas of specialization that warrant separate and detailed treatment. However, integrated systems that are provided by a manufacturer as a specific option are described.

PREHENSION MECHANISMS

Electric-powered prehension devices are available in a variety of forms, some of which resemble the anatomic hand while others do not and several of which are interchangeable. It is important to note that, appearances aside, all commercially available electric-powered prehension devices function in much the same way with a single degree of freedom of motion that brings two (or three) surfaces in opposition to allow for the grasping of objects. None of the devices offer independent movement of individual fingers, and all have fixed prehension patterns.

Early work on electric-powered prehension devices emphasized preservation of a handlike appearance. This preferential effort grew out of two broad, mutually reinforcing considerations. First, from the cultural vantage point, was a sensitivity to the sociological, symbolic, and aesthetic qualities associated with the human hand, qualities that can be powerful shapers of individual perceptions. The second consideration was a general expectation that in an environment of objects manufactured to be handled by human hands, a device with handlike characteristics would offer the best prehension function, an expectation that was taken literally with the adoption of shape as a principal characteristic.

Both of these considerations are as valid today as they were in the early years of electric hand design. Although quality of appearance can vary considerably, the cosmetic function of a prehensor with a handlike shape continues to be a strong determinant of personal acceptance. In addition, the broad contact surfaces of the electric hand and frictional properties of the cosmetic glove offer good grasp and retention of held objects. (Other significant factors cited for the acceptance of handlike prehensors-higher prehension force, reduced operating effort, increased comfort associated with the absence of control harnessing in myoelectri-cally controlled prehensors, and prehension control independent of the position of the prehensor with respect to the body-would apply equally well to electric-powered prehensors without a handlike shape.)

Electric hand prehensors have not, however, proved to be the ideal prosthetic solution that early developments were thought to foreshadow. Over two decades of experience with commercial electric hands have underscored the technological limitations of the designs and the deficiencies in our understanding of the physiology of the human hand, especially with regard to control. Fidelity to a handlike shape entails engineering compromises that diminish not only the prehensile function but also the overall mechanical function of electric prehensors. The handlike shape and fixed orientation of the fingers make precise tasks difficult to perform-a special consideration of persons with bilateral amputations but also cited by persons with unilateral amputations. The capability for reorienting the electric hand is significantly limited because of the associated loss of the physiologic wrist for most persons with upper-limb amputations, and it cannot be compensated by changing the orientation of the fingers. The electric hand's size and shape can visually obstruct the object being grasped or the work area in general. Shape constraints have also limited the form and arrangement of structural frames and finger armatures, and these parts can be damaged by heavy use. The material (polyvinylchloride [PVC]) from which most cosmetic gloves have been made has not been very durable and is susceptible to staining from common dyes, inks, and other materials. Power is lost in compressing and stretching the cosmetic plastic forms and gloves enclosing the mechanisms, and this contributes to the degradation of overall performance.

As a result of these observations, there has been increasing recognition that handlike prehension devices are most useful if supplemented with other prosthetic devices that have characteristics not constrained by fidelity to a handlike shape. Use of body-powered prostheses with hook-type prehensors is frequently cited in association with use of electric-powered prostheses. Additionally, adaptors and tools that can be held within the electric hand and mechanical tools that can be interchanged with the hand prehensors are available. More recently, several electric-powered prehensors that do not have a handlike shape have been introduced commercially to be used alternatively with or in place of electric hands.

General Characteristics of Commercial Electric Prehensors

Otto Bock Orthopedic Industry, Inc., and Hugh Steeper (Roehampton, England), Limited, both manufacture adult prehension devices with a handlike shape and devices not shaped like the hand. Hosmer Dor-rance Corporation also manufactures an electric prehensor that does not have a handlike shape. Specific characteristics of these devices are presented in Table 6C-1 for handlike devices and in Table 6C-2 for devices that do not have a handlike appearance.

Both the Bock and Steeper handlike prehensors are configured for palmar prehension-the opposition of the distal palmar pad of the thumb with the distal palmar pads of the index and middle fingers-and only the thumb and these two fingers are driven. Of the prehension patterns identified by Schlesinger, Keller et al. determined that palmar prehension predominated in the holding of objects for use. The persistence of this configuration in prosthetic hand designs and its general acceptance over the years supports their observation.

To achieve the palmar prehension pattern, the fingers of both types of electric hands are fixed in slight flexion at positions approximating the interphalangeal joints. The resulting finger shape also creates a concave inner prehension surface that is useful for cylindrical grasp. Additionally the frictional properties of the entire surface of the cosmetic glove of the electric hand provides for fixation and stabilization of objects against surfaces or against the body.

The prehension patterns of prehensors that do not have a handlike shape (those in Table 6C-2) are considered in the separate sections describing each type of device.

Table 6C-1 and Table 6C-2 list a variety of mechanical characteristics. Of these, the maximum prehension force, the maximum width of opening, and the speed of movement of the fingers merit some discussion because of their impact on the prehensile function of the devices.

Prehension Force

Force is a relatively easy characteristic to quantify. Therefore, it is often cited as a "figure of merit" for a prehension device. However, little is known about how prehension force capacity, frictional properties of the surfaces in contact, and conformability to surface features contribute to adequate grip. It is generally recognized that changes in either of the latter two characteristics can significantly alter the effectiveness of the applied force. Force should not, then, be considered in isolation from the other prehensile characteristics when drawing comparisons between particular devices.

Rationales for force requirements of prosthetic pre-hensors are typically based on physiologic performance. A study, done at the University of California, Los Angeles (UCLA), of human prehension force indicated that adult males could produce maximum mean forces of 95.6 newtons (21.5 pounds-force [lbf]) for palmar prehension, 103 newtons (23.2 lbf) for lateral prehension, 93.4 newtons (21.0 lbf) for tip prehension, and 400 newtons (90 lbf) for cylindrical grasp. More recent studies of larger populations have produced slightly different means but generally support the UCLA results.

Studies of forces applied in holding objects with the physiologic hand using palmar prehension and opposition of the thumb and index finger have shown that the static holding force is approximately one to two times the weight of the held object for objects weighing up to 3.52 kg (7.75 lb). The multiplication factor was found to depend on the friction between the skin and the surface material of the object held: the lower the coefficient of friction, the greater the force needed to hold the object. These results suggest an upper limit (based on applied force and frictional properties) on external forces against which human grip can be maintained without slippage. Using the UCLA data for maximum palmar prehension force, one could expect that the maximum force acting on an object that could be held without slipping could not exceed 47.8 newtons (10.8 lbf) for a low coefficient of friction and 95.6 new-tons (21.5 lbf) for a high coefficient of friction.

An unpublished investigation at UCLA indicated that prehension forces to a maximum of 66.7 newtons (15 lbf) were necessary to carry out a variety of activities of daily living. Peizer et al., reasoning that higher forces could only improve the prehensile utility of a prosthetic prehensor, proposed that this be a minimum standard for the maximum prehension force of an electric prehensor.

All of the devices in Table 6C-1 and Table 6C-2 have specified maximum prehension forces, some of them approaching or exceeding physiologic levels. However, the frictional properties of the materials lining the prehension surfaces and the ability of these materials to conform to the surfaces of held objects have not been specified. Consequently, one should be cautious in leaping to the conclusion that devices capable of achieving higher prehension forces can apply that force as effectively or more effectively than a device with a lower maximum prehension force.

Regulation of the applied force below the maximum is a function of the control system of the particular device and is discussed in the component sections. It should be noted that no commercial system provides direct sensory feedback of applied force and that force must be estimated indirectly through its effect on the object being grasped or the response of the prehensor as force increases.

All of the electric prehensors include some mechanism for maintaining the applied force in the absence of a control signal and without additional power to the motor, similar to the function of a vise. This is an important feature, essential to the overall performance of a prehensor. Without such a mechanism, it would be necessary to continue to drive the motor in stall to hold an object. During stall, a motor draws high currents, which would deplete a battery supply within a relatively short time.

The same mechanism that maintains the applied force also prevents the fingers from being pried open by external forces while an object is grasped. This feature is certainly helpful when using tools and other implements held in the prehensor. For safety and to prevent damage from excessive forces, all of the prehensors incorporate some method for opening the fingers when, for one reason or another, the prehensor does not respond to an opening control signal.

Width of Opening

In the handling of common objects, Keller et al.determined that 5.1 cm (2.0 in.) of prehensile opening was needed most of the time, but that an 8.2-cm (3.25-in.) opening was occasionally needed. Peizer et al.suggested the 8.2-cm (3.25-in.) opening as a minimum opening, a suggestion that was adopted by the Panel on Upper-Extremity Prosthetics of the National Research Council. Experience by users of prosthetic prehensors with an opening of 11.43 cm (4.5 in.) indicated a preference for the wider opening, although it was not used often. The maximum opening of any prehensor in Table 6C-1 or Table 6C-2 does not exceed 10.2 cm (4 in.)

Speed of Movement

Based on a study of user's experiences with electric prehensors available at the time, Peizer et al. recommended a minimum closure rate of 8.25 cm/sec (3.25 in./sec), measured at the fingertip. This minimum standard, considered a "high standard" in 1969, is exceeded by all of the prehensors in Table 6C-1 and Table 6C-2.

Data on physiologic finger speeds from an unpublished study at Northwestern University indicate maximum human finger velocities of approximately 40 radians/sec (2290 degrees/sec) for movements through a range of 75 degrees. Assuming a finger length from the metacarpophalangeal joint to the tip of 10 cm (3.9 in.), the maximum velocity at the finger tip would be 400 cm/sec (157 in./sec). These data provide an appreciation for the upper limit on physiologic finger speed, which is far in excess of the speeds attainable by any of the prosthetic prehensors.

In the same study, finger velocities were measured for an untimed pick-and-place task involving blocks of various sizes. Average finger velocities in this functional activity were considerably less than the maximum and were on the order of 3.0 radians/sec (172 degrees/sec). Only the Synergetic Prehensor, in Table 6C-2, achieves this speed, which was one of its design criteria. The efficacy of having prosthetic finger speeds on the order of functional physiologic speeds is greatly dependent on the control scheme with which the prehensor is operated. At higher speeds, a proportional relationship between the magnitude of the control signal and the response of the prehensor appears necessary to achieve confident and acceptable operation.

Otto Bock System Electric Hands (Adult Size)

The Otto Bock System Electric Hand is the most common type of electric hand prehensor used in North America. It is available in three adult sizes denoted by the circumferential dimension (in inches) at the knuckles. The 7 1/4 electric hand is suggested for adult females and juvenile males. The 7 3/4 size and 8 size hands are designated as being for adult males. The mechanism for all three hand prehensors, shown in Fig 6C-1.,B, is the same regardless of size. Different sizes are determined by the dimensions of a plastic hand-shaped form that is pulled over the skeletal mechanism, as in Fig 6C-1.,A. Gender differences and cosmetic coloration are provided by a separate glove made of PVC that is pulled over the plastic form. As described by Nader, each hand prehensor is therefore composed of three separate parts: the inner mechanism, a handlike form, and a cosmetic glove.

The hand mechanism (shown in Fig 6C-1.,B) includes the electric motor (mounted in line with the long axis of the arm), an automatic gear transmission, a support structure, and the finger assembly. Only the thumb and index and middle fingers are part of the mechanism and are oriented to provide palmar prehension. The motor drives the fingers (as one unit) and the thumb simultaneously in a plane perpendicular to the axis of the finger joints. The plastic form added over the mechanism incorporates the smaller two fingers. A wire frame within the form links these fingers to the middle finger so that they move somewhat in concert with the mechanized fingers.

When the fingers are in motion (i.e., not gripping an object), the transmission is in high gear, which allows the fingers and thumb to move at the speed noted in Table 6C-1. When an object is grasped, the transmission remains in high gear until the prehension force reaches 15 newtons (3.4 lbf), at which point it will automatically downshift to drive the fingers slower but at higher torque to a maximum prehension force of about 80 newtons (18 lbf). Without this automatic transmission, it would not be possible to achieve both the speed and maximum prehension force of the Bock hand with a single-motor design. In general, single-motor drive units are limited by a trade-off between speed and torque (the higher the speed, the lower the torque). The elegance of the Bock transmission is not without compromise. When an object is gripped tightly, it is not possible to release it immediately because the transmission must reduce the prehension force while in low gear until it reaches the lower limit, at which point it can shift to high gear and open the fingers.

The drive mechanism also includes a back-lock feature to maintain the prehension force when the motor is off and to prevent the fingers from opening. It is, however, possible to override the effect of the back lock, if necessary, by levering the hand to create very high forces at the fingertips and exceed the torque setting of a slip clutch. Operation of the slip clutch does not damage the mechanism, and the fingers can be closed manually, for appearance, until the control problem is corrected.

Although the back lock obviates the need to power the motor to maintain the prehension force, it is still possible to drive certain System Electric Hands in stall. Users of these particular prehensors must be advised to avoid this condition to conserve battery power. Bock has recently introduced an "energy-saving" design (in addition to their existing models) that senses the motor current and automatically cuts off power to the motor when a stall condition exists.

As previously mentioned, the primary prehension pattern of the System Electric Hand is palmar prehension. The mechanical arrangement of the thumb and fingers also provides cylindrical grasp for objects of moderate dimensions. For very wide objects (near the limit of the hand opening), the fingers are not able to encircle the object to secure it, but the plastic of the cosmetic glove provides friction that maintains a reasonably effective grasp. The use of a pliable hand form over the mechanism also improves the grasp, in general, because the inner surfaces of the hand are able to accommodate to the shape of objects, thus giving many points of contact between the prehension surfaces and the object being grasped.

Several options are available from Otto Bock to supplement the prehension features of the System Electric Hand. A pincer, or tweezer, that is keyed to fit the fingers of the hand prehensor can provide tip prehension for handling small objects. Alternatively, if the prehensor is equipped with a quick-disconnect wrist, it can be removed, and one of a variety of Bock work tools can be connected to the wrist of the forearm for special functions. These tools are not electrically powered. One could also exchange the hand prehensor for a System Electric Greifer (or Steeper Powered Gripper if the control system is compatible).

A variety of techniques are available for controlling the System Electric Hand alone. Otto Bock provides three myoelectric controllers as well as switch control. Bock terms the three myoelectric systems "digital two site," "grip force," and "double channel." The "digital two site" system is a two-site, two-function threshold myoelectric controller. When one or the other muscle site generates a myoelectric signal that exceeds the threshold, a control signal is generated to open or to close the prehensor. For the duration that the amplitude of the myoelectric signal is above the threshold, the prehensor will continue to operate, but the degree to which the signal exceeds the threshold does not alter the action of the mechanism. Regardless of the strength of the contraction generating the signal, the prehensor will move at only one speed or generate grip force (in low gear) at only one rate. In effect, the myoelectric signal is activating an electronic switch, and for this reason this type of control has been termed "myoswitch" control.

The "grip force" control is also a two-site, two-function system and is a variation on the "digital two site" controller that provides for two thresholds during closing. If the myoelectric signal from the "closing" site is above the lower threshold but below the higher threshold, the prehensor will close and apply force up to a maximum of about 15 newtons (3.4 lbf). To exceed this level, the user generates a stronger myoelectric signal, above the higher threshold, to cause the automatic transmission to downshift and the prehensor to apply higher force. Although two thresholds are involved, the controller is of the myoswitch variety, and the user cannot alter the speed of motion or the rate at which force is increased by varying the myoelectric signal. The higher threshold only provides a means of "manually" shifting the transmission.

The third Bock myoelectric system, "double channel," is a one-site, two-function myoswitch controller. The myoelectric signal from one muscle controls both opening and closing of the prehensor, depending on its amplitude with respect to one of two thresholds. When the signal is above the lower threshold but below the higher threshold, the prehensor closes. When the signal is above the higher threshold, the prehensor opens.

To be completely accurate in describing either the "grip force" or "double channel" system, it must be noted that the rate at which the myoelectric signal is generated is also important. To effect the function associated with the second threshold of either system, the amplitude of the myoelectric signal must not only exceed the threshold, but it must also do so within a short period of time. This is a subtlety of the decision process of the electronic controller that will not be elaborated here except to note that the lower-threshold function is generally associated with slower lighter contractions of the controlling muscle and the higher threshold is associated with faster and more forceful contractions.

For switch control, Otto Bock provides several types of electromechanical switches, including a cable pull switch, a harness pull switch, and a rocker switch. All switches provide operational positions for both opening and closing the prehensor.

Other manufacturers of myoelectric controllers have interfaced their systems with a special version of the Bock System Electric Hand that contains no electronics. Both Hosmer Dorrance Corporation and Motion Control provide two-site, two-function proportional myoelectric controllers. These controllers are notable because they enable the user to regulate the action of the prehensor (the speed of motion or rate of force application) in proportion to the amplitude of the myoelectric signal. Thus, lower signal levels produce slower movements or lower rates of prehension force application, and higher signal levels produce faster movements or faster rates of force application. Universal Artificial Limb Co. also has a two-source, two-function "variable speed controller" that can be used not only for proportional myoelectric control but also for control proportional to the output of variable position or force transducers.

Steeper Electric Hands (Adult Size)

Two adult-sized Electric Hand prehensors are available from Hugh Steeper and are denoted by their width across the knuckles: the 3-in. and 3 1/4-in. Electric Hands. For comparison to the Otto Bock adult hand prehensors, the 3-inch Steeper hand prehensor has a circumference at the knuckles of 7 3/8 in. and the 3 1/4 in. hand prehensor has a circumference of 7 3/4 in. The size of the prehensors are determined by the dimensions of the fingers and the enclosures around the drive mechanism. The enclosure is a two-piece hard plastic shell without fingers that gives the prehensor its handlike shape proximal to the finger joint (see Fig 6C-1.,A). The thumb, index, and middle fingers are molded of hard plastic directly over the armature of the finger assembly and are separate from the shell. The smaller two fingers are molded of pliable plastic and are attached to the plastic shell. For finishing, the shell and fingers are covered by a PVC cosmetic glove.

The mechanism, identical for the two sizes, includes a single motor with a gear reducer and drive screw and nut actuator, all held within a support structure (Fig 6C-1.,B). The first two fingers (as one unit) and thumb are linked to the nut and to the stationary support structure. As the nut travels along the screw, the fingers and thumb pivot and move in a palmar prehension pattern in a plane perpendicular to the joint axes of the fingers (the same prehension arrangement as used in the Otto Bock hand prehensors). The fingers move at a speed only slightly less than that of the Otto Bock hand prehensors. However, constrained by the trade-off between speed and torque of a single-motor design and lacking an automatic transmission like that used in the Bock mechanism, the Steeper Electric Hands achieve less than half the maximum prehension force of the Bock design.

A back-lock feature is inherent in the design of the drive screw and nut actuator, and the fingers cannot be forced open in typical usage when the prehensor is not powered. For safety purposes and to prevent damage to the mechanism under excessive forces, the thumb incorporates a breakaway device that allows it to hyperex-tend. Operation of the breakaway does not damage the thumb, and it can be manually reset to its normal position.

All models of the Steeper Electric Hands include current sensors that prevent the motor from running in a stall condition that draws high currents. Therefore, users need not consciously monitor their application of prehension force (when handling non-fragile objects) and are prevented from prematurely depleting their batteries. Additionally, a microswitch cuts off motor current when the Electric Hand is opened to its full extent, thereby preventing the motor from running in a stall condition in opening.

The thumb and finger arrangement of the Steeper Electric Hands provide both palmar prehension and cylindrical prehension. The PVC glove adds to the effectiveness of the prehension force by increasing the friction between the prehensor and the object being held; however, because of the hardness of the hand shell and molded fingers, the prehension surfaces cannot conform to the shape of the held object.

With respect to control, Steeper has two models of their Electric Hands: the "Myoelectric Hand" and the "Servo Hand." The Myoelectric Hand is somewhat of a misnomer since the prehensor model can be operated by means other than myoelectric signals. The Steeper controllers for the Myoelectric Hand include several parts: one or two transducers, a Digital Connector Ring (external to the hand prehensor), and an electronic assembly (in the prehensor itself). All of the control configurations operate the Myoelectric Hands in a switchlike manner, and the user cannot vary the speed of motion or the rate at which prehension force builds while generating the control signal.

Several types of transducers are available from Steeper. An Amplifier-Myoelectrode produces a control signal in response to a myoelectric signal that crosses an adjustable threshold. A Touch Activated Switch, similar in appearance to the Amplifier-Myoelectrode, produces a signal in response to a resistance change, such as from skin contact, between two of its metal surfaces. Two electromechanical switches are also available: the Momentary Contact Switch (similar to a push-button membrane switch) and a Single-Action Pull Switch, which is activated by a cable. All of these transducers operate from a single source and produce one control signal. To create a two-source, two-function controller, any two of these transducers are connected to the Digital Connector Ring. For example, one can have a two-site, two-function myoswitch controller using two of the Ampli-fier-Myoelectrodes. Alternatively, one can configure a two-source, two-function hybrid controller by using one Amplifier-Myoelectrode over an available muscle site to provide one function and a Single-Action Pull Switch operated by joint motion to provide the second function. Any combination is possible.

For situations where a single control source is all that is available, Steeper offers two options. The first is the One-Action Two-Function Adapter. This adapter accepts input from any one of the four transducers to provide one function, and whenever that control signal is absent, the adapter itself provides the second function automatically. This type of arrangement can be used to provide voluntary opening with automatic closing (up to the maximum prehension force) or voluntary closing with automatic opening (to full opening and activation of the limiting microswitch in the prehensor). The second option is the One-Muscle Two-Threshold Control, and as indicated by the name, it can be applied only if a myoelectric site is available. This single-site, two-function controller produces a "closing" command when the myoelectric signal crosses the lower of two thresholds and an "opening" command when the signal crosses the higher threshold.

The University of New Brunswick (UNB) single-site, three-state (single-site, two-function) controller can be used as an alternative to the Steeper single-site, two-function controller. The UNB controller can be adjusted more specifically to the characteristics of an individual's myoelectric control signal, a feature that may be helpful if the myosignal is marginal.

UNB also offers a two-site, two-function myoswitch controller compatible with the Steeper Myoelectric Hands and a single-site, single-function myoswitch controller that provides voluntary-opening control with automatic closing.

The second type of Electric Hand, the Steeper Servo Hand, offers a control method unique among commercial electric components. With this controller, the opening of the hand prehensor is determined by the degree to which a cable attached to a position transducer is pulled. (Fig 6C-2.). The further the cable is pulled, the more the prehensor opens, and the more the cable is slackened (and retracted by a spring in the transducer), the more the prehensor closes. The opening of the prehensor is therefore proportional to the displacement of the cable attached to the transducer, with full opening corresponding to about 9 mm (3/8 in.) of cable displacement. When the fingers close on an object, the force automatically increases to a maximum of 25 newtons (5.7 lbf) before the current sensor cuts off the motor.

It is important to note that although a cable is used to position the fingers, this control technique is not like an electric-power-assisted version of a body-powered voluntary-opening prehensor. In the case of a body-powered prehensor, the user has a direct sense through the control cable of not only the position of the prehensor's finger but also of its speed of movement and the force (inversely) exerted by it. With the Steeper servo hand, the user is linked by cable only to the transducer, which is remote from the prehensor and linked electrically to it. Therefore, the user directly perceives only the action of the transducer and force exerted on it, and not the action of the fingers and forces exerted by them.

Otto Bock System Electric Greifer

The System Electric Greifer, shown in Fig 6C-3., was developed by Otto Bock as an alternative to the System Electric Hand in work situations that require higher prehension force or that might damage the mechanism of the hand prehensor or damage or discolor the cosmetic glove. It can be easily interchanged for the hand prehensor when used with the Bock quick-disconnect wrist unit. The Greifer is available in one size and can be either a right or left unit. The mechanism is encased in a multiple-piece shell made of a durable hard plastic and is available with or without rubber pads lining the prehension surfaces of the fingers.

The Greifer's two fingers are broad surfaced and arranged to move symmetrically in opposition. They are articulated so that as they move, the distal prehension surfaces remain parallel to one another. The shape and articulation of the fingers provide lateral prehension and, for moderate-sized objects, cylindrical prehension. Adjustable tips, with or without rubber lining, provide tip prehension for handling smaller objects. The tips can be replaced with optional blanks machined for specific applications. (A screwdriver is required to adjust the position of the tips or to interchange them.)

In comparison to the System Electric Hand, the Greifer is longer by 3 cm (1.25 in.) and slightly heavier and has about the same maximum width of finger opening. In terms of mechanical performance, the Greifer is slightly faster and can develop significantly higher prehension force, 50% or 75% greater depending on the Greifer model.

The Greifer also incorporates an automatic transmission to enable the fingers to move relatively fast through space but to exert high forces when closed on an object. When the Greifer first closes on an object, it will grip up to a maximum force of 15 newtons (3.4 lbf), after which the transmission will downshift for gripping at higher forces. The short delay before the transmission downshifts enables users to grasp lighter and more delicate objects at the lower force and cease the "closing" signal before higher forces are applied. The transmission of the Greifer differs from that of the System Electric Hand in that there is very little delay between an "open" command and movement of the fingers even after high prehension forces have been applied.

As with the System Electric Hand, a back-lock mechanism prevents the Greifer's fingers from opening when power is not applied. For safety, this feature can be circumvented by one of two ways if the Greifer is not responding to an "opening" command. First, an external control wheel, in line with the motor, can be manually turned to drive the fingers open. (This control wheel also provides visual feedback of the action of the Greifer's drive mechanism during normal operation.) Second, a lever near the base of the fingers disengages the fingers from the drive transmission which enables them to be moved freely. Neither method damages the Greifer in any way.

In addition to the same wrist rotation capability of the System Electric Hand, the Greifer has built-in wrist flexion. The plane of flexion is parallel to the plane of motion of the fingers, which is perpendicular to the opposing prehension surfaces.

The Greifer can be operated by any of the control schemes available from Otto Bock for the System Electric Hand-a necessary capability if the Greifer is to be used interchangeably with the hand prehensor. The exception is "grip force" control, which is inherent in the design of the Greifer and thus available in association with the other control arrangements. As is the case for the hand prehensor, the Otto Bock controls for the Greifer are all of the myoswitch variety.

A model of the Greifer without electronics is available and can be operated with the two-source proportional controllers from Motion Control, Hosmer Dorrance, and Universal Artificial Limb Co. Therefore, an interchangeable Greifer and System Electric Hand can be used with any one of these controllers.

Hosmer NU-VA Synergetic Prehensor

The NU-VA Synergetic Prehensor, shown in Fig 6C-3., was designed as an alternative to a hand prehensor and with speed and force characteristics approaching those of the physiologic hand. It was developed by the Prosthetics Research Laboratory of Northwestern University with the support of the Department of Veteran Affairs and is manufactured by Hosmer Dorrance Corporation. The performance objectives of the prehensor are achieved with a two-motor design utilizing the concept of synergy.

Separate motors and gear trains are used to drive the two opposing fingers such that one finger is driven at high speed but low torque and the other finger is driven at low speed but high torque. Therefore, in the act of grasping an object, the prehensor's fast finger can quickly close on the object and the high torque finger apply the force, as necessary, to secure the object. The synergetic design also permits immediate release of objects when an "open" signal is generated because the high torque and high-speed fingers are driven simultaneously. In keeping with the design objectives, the maximum speed of movement of the fast finger of the Synergetic Prehensor is approximately that of the average speed of functional physiologic finger movements, and the maximum prehension force applied at the tip of the high-torque finger is approximately that measured for palmar prehension of adult males.

The mechanism and support structure of the prehensor are encased in a two-piece plastic shell, and the fingers of the Synergetic Prehensor are the same removable hook-shaped fingers as developed for the body-powered APRL (Army Prosthetics Research Laboratory) Voluntary-Closing Hook. The hook-shaped fingers provide powered lateral and tip prehension and passive hook prehension. For objects of moderate diameter with respect to the size of the prehensor, the lyre-shaped contour of the opening between the fingers provides for cylindrical grasp. The fingers are lined with neoprene to achieve higher contact friction during grasping. Neoprene pads are also arrayed on the case to facilitate activities in which the prehensor body is used to hold objects in place against other objects or to exert pushing forces on objects.

The drive train of the fast finger incorporates a back-lock mechanism that prevents the high-torque finger from pushing the faster, lower-torque finger back as objects are grasped. The back-lock, as with other prehensor designs, also enables objects to be held without continued operation of the motors. Should the prehensor not respond to an "open" signal when closed on an object, the fingers can be opened by a safety breakaway when external forces on the fingers exceed 133 newtons (30 lbf). This mechanism can be manually reset, and its operation does not damage the prehensor.

In addition to near-physiologic speed and force, the synergetic design is also energy efficient. Once the fast finger closes on an object and ceases to move, its motor is electronically cut off; therefore it does not run in stall during the application of force by the high-torque finger. To close on an object and grasp it with a force (at the fingertips) of 75 newtons (17 lbf), the prehensor draws an average of 138 mA or about 1.2 W. With a 100-mAh (milliampere hour), rechargeable 9-V transistor-type battery, the prehensor can perform approximately 1,300 cycles of opening and then closing to 75 newtons prehension force on a single battery charge. Therefore, it is possible to use these relatively small readily available batteries for a full days' use of the Synergetic Prehensor.

If the Synergetic Prehensor is used in association with an Otto Bock System Electric Hand, it is not advisable to use the 9-V transistor-type battery because of the current draw of the System Electric Hand. The Otto Bock battery could be used for this arrangement; however, its lower voltage will reduce the speed and force characteristics of the Synergetic Prehensor. For fittings of this type, it is recommended that an array of six or seven rechargeable AA batteries be used to provide the voltage requirements of the Synergetic Prehensor and the current requirements of the System Electric Hand.

Control of the Synergetic Prehensor is best achieved with a proportional system because of the speed of response of the device. A two-site, two-function proportional myoelectric controller is available from Hosmer Dorrance. This controller differs somewhat from other proportional myoelectric controllers in that the myoelectric signal is not smoothed by filtering but is used to generate full-voltage pulses that increase in width and number in proportion to the amplitude of the myoelectric signal. By processing the muscle signal in this manner and using the mechanical smoothing inherent in the drive system, the time delay associated with electronic filtering is eliminated, and the stiction of the mechanism is overcome. These two factors contribute to the almost instantaneous response of the Synergetic Prehensor and the ability to have good control even at low signal levels.

The two-source, two-function variable-speed controller from Universal Artificial Limb Co. can also be used with the Synergetic Prehensor. As noted, this controller can accept signals from myoelectrodes or from position or force transducers. The characteristics of the method of signal processing may result in some difference in the response time of the prehensor.

Steeper Powered Gripper

As has been the driving force for the design of all non-hand prehensors, the Powered Gripper was developed by Hugh Steeper, Ltd., to address various functional deficiencies associated with existing hand prehensors, constrained by their handlike shape and appearance. Interchangeable with the Steeper Myo-Electric Hand, the Powered Gripper (Fig 6C-3.) weighs approximately 25% less, moves with about twice the finger speed, and generates 70% higher maximum prehension force than does the adult Electric Hand. The improved performance was achieved by the synergetic action of separate drive systems and a different geometric arrangement for each of the two fingers.

The Powered Gripper uses the concept of one finger driven at high speed but low torque and the other at high torque but low speed, similar to that used in the design of the Synergetic Prehensor; however, the effect is produced differently. In the Synergetic Prehensor, the two fingers are pivoted about the same axis but are driven with greatly different gear ratios to achieve their individual speed/torque characteristics. With the Steeper Gripper, the fast finger has a gear ratio similar to that of the gear ratio of the slow finger, but its axis of movement is considerably closer to its drive screw than is the axis of the slow finger. The effect of the different pivot locations is a fast finger that moves over eight times faster than the slower finger but a slower finger that can produce four times the force of the fast finger. In comparison to physiologic performance, the fast finger of the Powered Gripper moves at about 70% the average velocity measured for functional finger movements and generates a prehension force about 63% the maximum palmar prehension force of adult males.

The body and fingers of the Powered Gripper are metal castings. The fingers are contoured to provide passive hook prehension and have flattened opposing surfaces for powered lateral and tip prehension. The opening between the fingers is also contoured to accommodate cylindrical objects. The prehension force of the Powered Gripper is made more effective by the use of relatively soft frictional rubber pads to line the fingers. The pads are grooved over a portion of their surfaces, and the fingers are hollowed beneath the material. The grooves and hollowing allow the pads to deform and mold to the shapes of held objects, which distributes the prehension force over a broader contact area.

The mechanism of the powered griper includes a drive screw and nut assembly as the last stage in actuating the fingers. As with the Steeper Electric Hands, this assembly cannot be back-driven and therefore provides a back-lock feature keeping the fingers in place when unpowered. The drive screws for both fingers are connected to plastic wheels on the outside of the pre-hensor's case. These wheels can be turned manually to open the fingers in the event that the prehensor does not respond to an "open" command.

The same control schemes as used for the Steeper Myoelectric Hand are compatible with the Powered Gripper. (There is no Servo version of the Gripper.) In closing, the fingers operate sequentially. The fast finger first moves to establish contact with the object to be grasped and, at a force of 15 newtons (3.4 lbf), stalls. A 600-ms delay follows before the slow finger becomes active. Since the Steeper controller does not provide proportional control of motion, the delay gives the user time to cease the closing signal if a low-force grasp is wanted. If the closing signal is not interrupted during the 600-ms delay, the slow finger is activated, increasing the prehension force to 60 n ewtons (13.5 lbf), at which point the slow finger stalls. It is important to note that the Steeper electronics prevents the motors from running in a stalled condition. Therefore, only one motor is active at a time, and once the slow finger stalls, no additional motor current is drawn while an object is held. The opening sequence is the reverse, with the slow finger opening first (if it had been activated in closing), followed, without a delay, by the opening of the fast finger.

As mentioned, the control schemes for the Steeper Powered Gripper are switchlike, and the user cannot regulate the speed at which the fast finger moves or the rate at which the slow finger increases the force. This arrangement is compatible with the Otto Bock "digital two site" (two-site, two-function) myoswitch control and the Otto Bock electromechanical switch controllers. Versions of the Gripper are made with an Otto Bock quick-disconnect wrist for interchange with Otto Bock System Electric Hands using either of these control schemes.

NY-Hosmer Prehension Actuator

The NY-Hosmer Prehension Actuator (PA) is not of itself a prehension device and for that reason is not listed in Table 6C-2. The PA, shown in Fig 6C-4., is a motorized winch that provides electric-powered operation of the cable-actuated Hosmer Dorrance voluntary-opening split hooks. It was originally designed by William Lembeck of New York University as a complete forearm setup for use with a body-powered or electric-powered Hosmer Dorrance elbow. In that configuration, the mechanism occupies the distal 10.8 cm (4.25 in.) of the forearm with a rotation joint proximal to the mechanism. The forearm segment proximal to the rotation joint contains the forearm saddle assembly for the elbow and, because of the saddle's dimensions, has a minimum length from the elbow axis of 9.5 cm (3.75 in.). The complete forearm setup has a minimum length from the elbow axis to the distal face of the wrist of 20.3 cm (8.0 in.). Longer forearms are provided by lengthening the distal forearm segment, thus keeping the weight of the mechanism (about 218 g; 0.48 lb) as proximal as possible. Rotation to orient the split hook is done proximal to the PA in order to maintain an efficient alignment between the cable attachment post of the split hook and the cable leading from the actuator mechanism.

The PA is typically powered by a 6-V battery pack and, at that voltage, can open a split hook having four or five rubber bands. The time to open the hook to its limit is dependent on the number of bands used with the hook but is on the order of one second.

Operation of the PA is with a single-source controller. Hosmer offers a single-site, single-function myos-witch controller and a variety of single-function electromechanical switches. A three-function cable pull switch is also available for operation of the PA and the NY-Hosmer electric elbow by using one control action. (This is the configuration shown in Fig 6C-4.). The first two functions of the switch are for the elbow; the third is for the PA.

Controlled activation of the PA causes it to pull the split hook open. If the control signal stops before the PA pulls the hook to full opening, the hook is closed by the rubber bands. If the control signal is maintained after the hook reaches full opening, the PA is electrically cut off (so that it does not draw motor current), and the motor is dynamically braked. The dynamic braking, which is maintained as long as the control signal is present, allows the split hook to close but at a slow speed. This action gives the user time to adjust the position of the split hook relative to the object being grasped. When the control signal is withdrawn, the braking is removed, and the split hook closes freely.

Since the original introduction of the PA forearm setup, it has been adapted to below-elbow (transradial) fittings with single-site myoswitch control. To use this configuration, one must take into account the forearm length requirement of the PA and the placement of the PA battery pack. A transradial fitting of this sort can be done with a supracondylar self-suspending socket and does provide some advantage over fitting of an electric prehensor by shifting most of the weight proximal to the wrist.

Commentary on Electric Prehensors

The interplay of psychological and social aspects associated with the human hand and the need for prehension function and independent capability are complex. Generalizations favoring one type of prehensor over any other are limiting, and there is little consensus among users of prosthetic prehensors as to which device is best suited as a replacement for the physiologic hand. Even the similarity to the anatomic hand that is possible with the present-day Electric Handlike prehensors is not universally desired. Some persons, particularly those with bilateral amputations, are sensitive to the prehension and performance advantages of prehensors not having a handlike shape. Other persons, finding the apparent cosmesis of Electric Hands insufficient and being repelled by it, prefer a device that has a form "truer" to its gripping function. Until a more versatile anthropomorphic prehension device is developed, the need for a variety of options will remain.

WRIST MECHANISMS

Studies of persons using their hands to perform various common activities and occupational tasks have shown significant utilization of forearm rotation and wrist motions in the performance of these actions.Most of the activities studied revealed a range of motion through which the joint moved during the course of an activity as opposed to a variety of fixed positions across the activities. In studies having many different activities, the total range of motion spanned was found to be approximately 100 degrees for forearm rotation, 80 degrees for wrist flexion and extension, and 60 degrees for wrist radial and ulnar deviation. For the specific task of eating, the total range of motion was about 100 degrees for forearm rotation, 30 degrees for wrist flexion and extension, and 30 degrees for radial and ulnar deviation of the wrist.

Except for the Otto Bock Electric Wrist Rotator, all commercial prosthetic wrist components are purely mechanical. There are many factors that make development of electric-powered wrist components particularly difficult. From a component design viewpoint, there are the constraints of size and weight imposed by the location of the joint. The device must fit within a cylinder of about 5 cm (2 in.) in diameter and occupy as little length as possible so as to accommodate (ideally) a variety of residual-limb lengths.

The component must also be relatively lightweight to minimize counterforces exerted on the residual limb in the case of a transradial fitting or minimize the counter-torque that would reduce the lift capacity of a prosthetic elbow in higher-level fittings. And although lightweight, the structure of the component must be robust enough to withstand the forces exerted on the prehensor and transferred back to the residual limb through the wrist joint. Also a consideration with respect to weight is the need for relatively low power consumption to eliminate an additional battery if used in conjunction with other electric components.

With regard to function, there is the question of what joint motions should be provided. The anatomic forearm and wrist joints can be approximated by a triaxial joint with the axes of rotation, flexion, and deviation (roll, pitch, and yaw) having a point of intersection near the base of the prehensor. All three motions have been shown to contribute to functional activities.

There is the issue of control. At least one additional control source would be needed for each powered joint of wrist motion, unless the control system operates in a sequential manner. Even sequential control would require at least two control sources-one for selection and one for movement control.

Finally, the performance of the component must be exceptionally better than the alternatives for the user to be attracted to its operation. For the person with a unilateral amputation, the primary alternative is the intact limb, which can be preferentially used for activities involving significant forearm and wrist motion. On the prosthetic side, one could use compensatory motions of proximal physiologic joints and have manually positioned mechanical wrist components that would offer adjustable fixed orientations of the prehensor. Although operation of these components typically involves the physiologic hand, the operation is relatively quick and straightforward. (The fact that this technique is so widespread underscores the remarkable qualities of the physiologic wrist. Persons using this technique do so without giving much thought to what they are doing with their intact wrist and hand while using them to position the prosthetic wrist.)

For persons with bilateral arm amputations, there are alternative methods for actuating and positioning mechanical wrist components that do not necessarily require the contralateral limb. However, these components cannot, in general, be operated so as to perform work such as turning a handle. Neither can they typically be adjusted dynamically during a motion, such as adjusting the wrist attitude while raising a utensil to one's mouth. Although these deficiencies have inspired many designers to attempt a more versatile electric-powered wrist, advances have been slow to come, and no multiaxis components have been realized commercially.

Otto Bock Electric Wrist Rotator

The Electric Wrist Rotator developed by Otto Bock, shown in Fig 6C-5., addresses many of the difficulties outlined in the preceding section to provide the functional analogue of forearm rotation. The drive unit is a single motor with a gear reducer having a rotation axis in line with the longitudinal axis of the forearm. It is structurally supported within the lamination collar of the Bock quick-disconnect wrist and can fit any of the three sizes of wrist lamination collars, which have diameters of 4.0, 4.5, and 5.0 cm (1.6, 1.8, and 2.0 in.). Its length is 6.7 cm (2.6 in.) from the distal edge of the lamination collar to the proximal surface of the motor housing.

The rotator is relatively lightweight at 96 g (0.21 lb), approximately 20% of the weight of a Bock System Electric Hand. It is also relatively energy efficient and draws a no-load current of 150 mA. (The stall current is 700 mA.) The power requirements are such that it is feasible to operate a Bock System Electric Hand, or Greifer, and the Electric Rotator from a single 6-V Bock battery. However, the ability to get through a full day's use on one battery will vary according to the degree to which the devices are utilized.

The rotator mechanism is also protected from external forces through its attachment to the wrist lamination collar. Side forces and axial forces exerted on the prehensor are transferred to the lamination collar and prosthetic forearm rather than to the rotator mechanism. Excessive torques on the prehensor will cause the ratchet of the prehensor's portion of the quick-disconnect wrist to slip rather than back-drive the wrist mechanism.

The coaxial electrical coupling of the Bock quick-disconnect wrist allows the rotator to turn an electric prehensor continuously in either direction. In general, however, the performance characteristics of the rotator have been compromised to achieve the necessary size, weight, and power characteristics. The rotator does not generate high torque and cannot be used for work, e.g., turning valves or door handles, unless the resistance is minimal. The rotation is primarily for preposi-tioning and changing the orientation of the prehensor prior to an action or while the prehensor is holding a lightweight implement, such as a utensil with food or a cup of liquid. The speed of response is also compromised. At a no-load speed of 8.33 rpm (0.87 radians/sec or 50 degrees/sec), it is perceptibly slower than physiologic forearm rotations. Physiologic forearm rotation can achieve time-averaged maximum velocities in excess of 14 radians/sec (800 degrees/sec) for pronation and 20 radians/sec (1,150 degrees/sec) for supination.

Otto Bock provides two control schemes for the Electric Wrist Rotator: control by electromechanical switches or myoswitch control in conjunction with a Bock electric prehensor. The Bock two-function switches include a cable pull switch, a harness pull switch, and a rocker switch. The myoswitch control, termed "four channel control", is a two-site, four-function controller that operates both an electric prehensor and the wrist rotator. In this system, one muscle site controls one function of the prehensor (e.g., closing) and one function of the rotator (e.g., pronation) by using the magnitude and rate of contraction to distinguish the component to be controlled. The second muscle site controls the other function of each component, again using the magnitude and contraction rate to direct the control to the appropriate component. Both muscles are needed to control each of the components. As with all Bock control systems, the performance of the component is not influenced by the amplitude of the myosignal once the component is selected. The rotator will operate at a single speed.

Motion Control provides a two-site, four-function version of their proportional myoelectric controller to operate both an electric prehensor and the electric wrist rotator. This system channels the signals from both muscle sites to each component and uses cocon-traction of the agonist-antagonist pair to switch from prehensor to rotator and vice versa. The approach has the advantage, in comparison to the Bock two-site, four-function controller, of allowing the user to regulate the action of either selected component in proportion to the amplitude of the myosignal. In practice, the proportional control is of more obvious a benefit for operation of the prehensor. The speed of the rotator is such that users appear to operate it near its maximum even for small corrective actions.

The rotator can be independently controlled by the proportional two-site, two-function myoelectric controller from Hosmer Dorrance. It can also be controlled with the two-source, two-function "variable speed controller" from Universal Artificial Limb Co. As described in the section on prehensors, this controller can accept input from myoelectrodes or from force or position transducers.

Wrist Flexion Units

Although there are no commercial components that provide electric-powered wrist flexion, this is an important function for the person with bilateral arm amputations and for some persons with unilateral amputations. Therefore, it is useful to know how this function can be provided in prosthetic fittings involving electric prehensors.

The Otto Bock System Electric Greifer is unique among commercial prehensors because it incorporates a flexion joint within the prehensor. The joint is a manually positioned friction joint that can be adjusted for more or less friction. The range of motion is plus and minus 45 degrees and occurs in a plane perpendicular to the prehension surfaces of the fingers.

For other electric prehensors, engineers and pros-thetists have devised a variety of techniques for adapting commercially available mechanical flexion components for use with electric prehensors. The Sierra Wrist Flexion Unit, the Hosmer Flexion-Friction Wrist, and the United States Manufacturing Company (USMC) E-Z Flex Wrist have all been used in clinical electric-powered fittings. Modifications to both the wrist component and the wrist coupling of the prehensor may be required, depending on what specific components are being used together. The Hosmer Universal Shoulder Joint, with appropriate-sized proximal and distal lamination collars, has been adapted as a friction-type wrist flexion joint. In any of these configurations, the important considerations are to provide a pathway for the electrical wires to the prehensor and to limit, with a mechanical stop, any rotation joint crossed by the wires so that the wires will not be damaged by unrestricted continuous rotation in one direction.

A configuration developed to use a flexion wrist unit with the Bock Electric Wrist Rotator in a forearm setup used with a prosthetic elbow is shown in Fig 6C-6..The forearm is composed of two sections joined by a Bock quick-disconnect wrist. The proximal section contains the wrist rotator, positioned as close to the elbow as possible without interfering with the full flexion of the elbow. The distal section, essentially a hollow cylinder, incorporates a modified Bock quick-disconnect adaptor at the proximal end to mate with the wrist rotator and a flexion wrist unit at the distal end, which may require an adaptor to mate with a specific prehensor. The rotator is electrically operated. The flexion unit is manually positioned.

Commentary on Wrist Components

The significant participation of forearm and wrist motions to provide fine orienting and positioning of the human hand is well documented. The need for this capability in a prosthetic limb is no less great. It is even possible that the prosthetic wrist takes on more significance in the context of the total prosthesis. Prosthetic fingers cannot be repositioned within the prehensor to accommodate orientation needs as changes in the position of physiologic fingers can be made to complement the anatomic wrist position. And compensatory motions of proximal physiologic joints may be restricted by the suspension of the prosthesis or the harnessing for control actions. As one considers persons with bilateral amputations, especially persons with amputation levels above the elbow, the need for assisted wrist function, on at least one side, becomes even more demanding. The technological obstacles and control problems are severe. However, the potential functional advantages of better wrist components will likely continue to drive development efforts.

ENHANCEMENTS TO BODY-POWERED ELBOWS

Hugh Steeper, Limited, offers two optional electrical enhancements for its body-powered mechanical elbows. The first is the Steeper Interlock System, which is an electromechanical switch actuated by the locking mechanism of the elbow. The second is the Steeper Electric Elbow Lock, which is a motorized lock for an otherwise body-powered elbow.

Steeper Interlock System

The Interlock System was developed to allow single-cable control of a mechanical elbow and an electric-powered cable-actuated Steeper Servo Hand. In this configuration, the control cable is routed from the control harness, through a forearm flexion attachment, and to a termination on the position transducer (mounted distally in the forearm) that operates the Servo Hand. When the elbow is locked, the elbow interlock switch is in the "on" state, and pulling on the control cable operates the Servo Hand. When the elbow is unlocked, pulling on the control cable both flexes the elbow and actuates the transducer for the Servo Hand. However, the interlock switch within the elbow is in the "off state, and the Servo Hand is prevented from responding to the transducer actuation. Therefore, pulling on the control cable operates one or the other component depending on the state of the elbow lock.

While this arrangement is operationally similar to that of a body-powered elbow used with a body-powered, voluntary-opening split hook, forces used to flex the Steeper elbow (with interlock switch) do not alter the force of prehension of the Servo Hand. In a total body-powered prosthesis, the force in the control cable that flexes the elbow is also transferred to the split-hook prehensor and proportionally diminishes the prehension force exerted by the rubber bands. As the elbow flexion force is increased because of the weight of an object being actively lifted, a limit is eventually reached where the prehensor exerts no holding force. In a prosthesis with the Steeper Interlock System, the Servo Hand is electrically disconnected whenever the elbow is unlocked and free to move. The force of prehension remains constant regardless of how much force is exerted on the control cable to flex the elbow.

The Steeper mechanical elbow and Interlock System can be used in combination with other prehensors, such as the switch-controlled versions of the Bock System Electric Hand and Greifer. In this configuration, a Bock Harness Pull Switch is sewn into the control strap (with a shunt strap to protect the switch from high forces), and one of the battery leads for the Bock prehensor is connected through the interlock switch. When the elbow is locked and the interlock is in the "on" state, pulling on the harness extends the harness switch and operates the prehensor. When the elbow is unlocked and the interlock is in the "off" state, the harness switch extends during elbow flexion but has no effect on the prehensor.

It is important to note that the control configuration with the Steeper Interlock provides sequential control of the elbow and prehensor. It is not possible to operate both devices in a coordinated manner as is possible with independent-control hybrid configurations, such as a body-powered elbow and myoelectrically controlled prehensor using biceps and triceps muscles.

Steeper Electric Elbow Lock

The electric lock for the Steeper mechanical elbow is a straightforward alternative to the body-actuated mechanical lock. Just as with the mechanical elbow lock, each operation of the electric lock changes its state: from locked to unlocked or from unlocked to locked. The electric lock, powered by a 6-V battery, is operated by an electromechanical switch or by a single-site myo-switch control. Steeper offers a variety of switches, and any other commercially available switch that provides a momentary switch closure can be used as well. The variety of control arrangements possible with the electric lock is a major advantage for persons who could benefit from a cable-actuated mechanical elbow but who have difficulty producing the control motions or forces required by the mechanical lock.

ELBOW MECHANISMS

Three electric elbows are available for adults: the Boston Elbow, the NY-Hosmer Electric Elbow, and the Utah Arm. These elbows differ from one another in mechanical configuration, drive mechanism, and control options. Table 6C-3 summarizes various characteristics of these devices, and each of the elbows will be discussed individually in following sections.

In addition to the powered elbow joint, all of the elbows incorporate a friction joint, or turntable, for manual humeral rotation. With the Boston Elbow and the NY-Hosmer Electric Elbow, the friction is adjusted by a crown nut on a threaded stud centered in the proximal surface of the elbow enclosure. Access to this nut must be provided in the fabrication of the humeral shell. The Utah Arm utilizes an external split collar for friction adjustment of the humeral rotation joint; therefore, no special accommodation must be made in the fabrication of the humeral shell for access to the adjustment.

Much concern is given to the lifting capacity of electric elbows. While this is an important characteristic, especially for persons with bilateral amputations, the elbows are primarily used to position the prehension device and then kept in place while performing some activity. As noted in Table 6C-3, the three elbows have maximum live lift capacities ("live lift" meaning lifting by powering the elbow) of between 3.4 N-m and 5.9 N-m (2.5 ft-lb and 4.5 ft-lb). At a distance of 30 cm (approximately 12 in.) from the elbow axis, the NY-Hosmer Electric Elbow can lift a maximum weight of 1.1 kg (2.5 lb), and the Boston Elbow can lift a maximum weight of 2.0 kg (4.4 lb). The maximum live lifting weight for the Utah Arm lies between these two values. Any weight due to the materials of the forearm, wrist component, and prehension device must be subtracted from these values to arrive at an estimate of the maximum weight of an object that can be held and lifted. An adult Electric Handlike prehensor weighs on the order of 0.45 kg, or about 1 lb. Assuming an elbow axis to palm distance of 30 cm, having this type of prehensor would reduce the maximum weight of an object that can be lifted to approximately 0.65 kg (1.4 lb) for the NY-Hosmer Electric Elbow and to 1.55 kg (3.4 lb) for the Boston Elbow. Weight of the forearm and wrist componentry would further reduce these values.

In comparison, the lifting capacity of the physiologic elbow can exceed 25 kg (55 lb) for an adult male at low speeds of flexion and over 13 kg (29 lb) at flexion speeds of about 57 degrees/sec. Therefore, one cannot expect to perform the same types of activities, especially those involving the active lifting of moderate to heavy loads, with an electric elbow as one would expect to do with the physiologic elbow.

Heavier loads can be lifted by a prosthesis with an electric elbow, but in a passive manner. This is done by locking the elbow in place after prepositioning it, using body movement and posture to orient the prehensor to grasp the object, and then straightening the body without actively moving the elbow joint. In this way, objects can be lifted that exceed the live lift capacity of the elbow. However, even this technique is limited by the breakaway device or slip clutches incorporated in the elbow mechanisms to protect them against mechanical overload. This overload protection also serves to protect the user, to a degree, from excessive forces transferred through the socket during accidents such as falls. The elbow would give way if the person fell upon the prosthesis. Both the Boston Elbow and the Utah Arm have passive lift capacities of 68 N-m (50 ft-lb), and the NY-Hosmer Electric Elbow has a capacity between 24.4 N-m and 27.1 N-m (18 to 20 ft-lb). By using the prosthesis configuration described earlier-with an electric prehensor and distance to the elbow axis of 30 cm (12 in.), the Boston Elbow and Utah Arm can passively lift an object weighing up to 23 kg (49 lb), and the NY-Hosmer Electric Elbow can passively lift 8.1 to 9.1 kg (17 to 19 1b).

As with lift capacity, the speed of elbow motion is often used as a figure of significance when comparing prosthetic elbows. But here again, some perspective can be gained by considering speeds of electric elbows in comparison with physiologic performance. Averaged maximal speed of physiologic elbow flexion for adult males has been measured at about 600 degrees/sec for movements through a 120-degree range, with peak speeds in excess of 900 degrees/sec. Clearly the maximum speeds of adult electric elbows (see Table 6C-3) are far less than these values. However, maximum speeds of elbow flexion are probably rarely used in everyday functional activities. Peak physiologic elbow speeds more typical of those that might be used in common functional activities have been found to be correlated to the amplitude of the movement with the approximate relationship: speed (degrees/sec) = 2.9 degrees/sec/degrees x distance (degrees). For a movement over a 10-degree range, the peak velocity during the movement would be about 29 degrees/sec. For a greater angular movement of 90 degrees, the peak velocity would be about 261 degrees/sec. Therefore, it would appear that all of the electric elbows can approach functional physiologic speeds over short distance movements but are significantly slower than the physiologic elbow over larger angular movements.

Perhaps more important than an electric elbow's measured speed is how it is being controlled in relation to its speed of response. To use an extreme example, it would be difficult to position a fast elbow by using switch control that actuated the elbow at full speed in flexion and in extension. The user would have a tendency to overshoot the target position and would likely not be able to make small changes in position. Therefore, as electric elbows have become faster, there has been greater utilization of proportional velocity control. In this type of control, the magnitude of the input signal, which the user is presumed to be able to regulate, determines in direct proportion the speed of motion. By creating a higher-amplitude signal, the user directs the elbow to move faster (up to the limits of the mechanism), and by producing lower-level signals, the user drives the elbow at slower speed.

The following sections elaborate on each of the adult electric-powered elbow systems.

Liberty Mutual Boston Elbow

The Boston Elbow had its origins in the 1960s in a cooperative research and development venture involving the Liberty Mutual Insurance Company and its Research Center, the Massachusetts Institute of Technology, the Harvard University Medical School, and Massachusetts General Hospital. The first prototypes were encouraging, but considerable development by the Liberty Mutual Research Center during the first 5 years of the 1970s was necessary to produce a version that could be commercialized. Robert Jerard redesigned the original prototype and proved that a commercial version was feasible, and T. Walley Williams III carried out the commercialization and directed subsequent design alterations. Trials with the commercial elbow were begun in 1975, and the elbow was made generally available in 1979.

In its present form (see Fig 6C-7.), the Boston Elbow is available in one size and is configured with the motor and gearing within the elbow "cap" and the battery and electronics supported in a metal forearm frame. A prefabricated plastic and foam forearm shell (not shown in the figure) is custom-shaped and laminated to enclose and protect the forearm componentry when the prosthesis is finished.

The elbow is typically controlled in one of two ways: by a two-site, two-function myoelectric controller or by integrated or separate two-function electromechanical switches. The two-site myoelectric controller offers control of speed and torque in proportion to the magnitude of the myoelectric signal. Separate gain adjustments allow for tailoring the response of the elbow with respect to the condition of the myoelectric sources. An Evaluation Meter is available to monitor the myoelectric signals for evaluation and training and can be used with or without the elbow in operation.

Control by electromechanical switches provides single-speed (or single-torque) operation in flexion and in extension. The additional circuit board required for switch control includes circuitry for limiting the flexion and extension speeds separately. Although the user cannot vary the speed of motion when using switches, the speed can be set to an acceptable level. These adjustments can be helpful during training, when the speeds might be reduced while the client becomes familiar with the operation of the elbow. The speed adjustments are also helpful in balancing the elbow's response to gravity, which (as with all elbow mechanisms) assists the elbow during extension and retards it during flexion, especially if an electric prehensor and electric wrist rotator are used.

In operation, the Boston Elbow can be positioned anywhere within its 135-degree range of motion and is self-locking whenever the control signal ceases. A free-swing range of 30 degrees of flexion from the stopped position of the elbow can be engaged and disengaged by manual operation of a mechanical slide bar.

The Boston Elbow can be used in conjunction with cable-actuated body-powered prehensors and with other electric components in a configuration with separate control sources for each component. Use of separate control sources is preferred, when feasible, because it can allow for simultaneous and coordinated operation of more than one component. However, if control sources are limited, Liberty Mutual offers a two-site, four-function proportional myoelectric controller configured for sequential operation of the elbow and an electric prehension device. A separate switch, such as a harness-type switch, is used to select the component to be controlled. Although three control sources are needed-two myoelectric sites to control the movement of the component and a source to actuate the selection switch-this arrangement provides for proportional control of each of the two components. Other customized control configurations have been developed by the Liberty Mutual Research Center, and circuit diagrams for the elbow controller are readily available.

NY-Hosmer Electric Elbow

The NY-Hosmer Electric Elbow was designed by William Lembeck at New York University under the direction of Sidney Fishman. The prototype of this mechanism was originally conceived for use by children and was evaluated, as such, in the 1970s. Consequent modification to that prototype design and the involvement of the Hosmer Dorrance Corporation resulted in the commercialization of "large" and "medium"-sized versions introduced in 1983. The two sizes are equivalent to the E-400 and E-200 Hosmer Dorrance mechanical elbows, and the mechanical elbows can be alternatively fit to prostheses originally configured with the electric elbow. Hosmer Dorrance also introduced versions of the elbows for exoskeletal and endoskeletal applications (Fig 6C-8).

The same motor and drive mechanism, contained in the elbow cap, is used for all versions of the elbow; therefore, mechanical performance characteristics are the same for all models. External dimensions, the turntable, and the forearm saddle attachments vary from model to model. The absence of fixed componentry in the forearm and the use of a forearm saddle provides considerable freedom in the length and customized shaping of the forearm section.

The elbow is powered by a separate battery pack, available in four-and five-AA cell configurations, that can be positioned within the prosthesis as appropriate. Placement within the humeral section is preferable to placement in the forearm because additional weight in the forearm will reduce the functional lift capacity- the maximum weight of a held object that can be lifted by the elbow.

Two control options are available from Hosmer Dor-rance: switch control using two-function electromechanical switches and two-site, two-function myoswitch control. A variety of electromechanical switches are available from the manufacturer, including cable and harness pull switches and one-site and two-site push switches. Other switch configurations are also possible. Both the switch control and the myoswitch control operate the elbow at one speed, which cannot be adjusted but which is determined by the battery voltage, the load on the elbow, and the direction of movement.

The variable-speed controller manufactured by Universal Artificial Limb Co. has been adapted to the NY-Hosmer Electric Elbow. This two-source controller can be set up to accept input from force-sensitive pads, displacement transducers, or Otto Bock electrodes to provide proportional control of the elbow's speed.

A pawl-type locking mechanism placed in an early stage of the drive train locks the elbow virtually anywhere through its 130-degree range. Locking is automatic whenever the control signal ceases. The elbow can also be made to swing freely by driving it to its fully extended position, at which point the free swing automatically engages. Once engaged, the elbow can be swung or pushed unpowered anywhere within its full range of motion. Free swing is disengaged by activation of the flexion control. Elbows can be equipped with or without the free-swing feature, and elbows without free swing can be retrofit to incorporate it.

Numerous configurations are possible when the electric elbow is used in conjunction with wrist and prehension components having control sources separate from the source (or sources) operating the elbow. Hosmer Dorrance does not offer methods for integrating control of the elbow with other electric components, with the exception of the NY-Hosmer Prehension Actuator (PA). In configuration with the PA (shown in Fig 6C-4.), the elbow can be operated by a three-function cable pull switch. The first two functions operate the elbow in flexion and extension, and the third function (with the switch control cable fully extended) operates the opening of the PA.

Motion Control Utah Arm

The Utah Artificial Arm is manufactured and distributed by Motion Control, a division of IOMED, Inc. The system was developed at the University of Utah in the latter half of the 1970s by a team directed by Stephen Jacobsen, Ph.D. The original Utah Arm, as first clinically fit in 1980, included the electric elbow mechanism and control electronics developed by the Utah team and a body-powered voluntary-opening split-hook prehensor. In 1982, Motion Control introduced a proportional myoelectric controller that allowed the elbow to be used in conjunction with an electric prehensor. As it is presently configured, the "arm" includes a motorized elbow mechanism, a friction-type humeral turntable, a forearm shell, and electronics for both the elbow and an optional electric prehension device (either an Otto Bock System Electric Hand or Grei-fer).

The Utah Arm is available in one size and is shown in Fig 6C-9. The battery pack and elbow electronics are contained within the stationary (with respect to the humeral section) enclosure distal to the turntable. The motor, mechanical transmission, and prehensor electronics are located in the forearm section. The forearm shell is a finished injection-molded plastic enclosure that can be cut to shorter length or lengthened by the addition of an extension. Elbow rotation occurs about an axis through the anterior aspect of the joint. This placement allows flexion to approximately 150 degrees, thus bringing the prehensor nearer to the face with less shoulder flexion than is possible with other elbow designs. Modularity of the electrical and mechanical assemblies is a hallmark of the Utah Arm, and this facilitates access for troubleshooting and replacement of subunits.

A single control technique is used to operate the elbow mechanism: two-site proportional myoelectric control. Switch control is not feasible because of the relatively high speed of the elbow, over 100 degrees/sec with an electric prehension device. Nonlinear filtering of the myoelectric signals provides for quick response of the elbow to sudden high-amplitude changes in the control signals to achieve fast movements, but smoother response for the slower-changing lower-amplitude signals used in more precise movements. Motion Control offers the MYOLAB II-EMG Tester/ Trainer, which incorporates meters and auditory feedback of myoelectric signal amplitude for evaluation and training. The MYOLAB II can be used to monitor the myoelectric signals simultaneously with operation of the Utah Arm.

Locking of the elbow is engaged whenever it is held stationary for a set period of time (the length of which can be adjusted) or whenever a momentary switch is actuated. The elbow has 22 locked positions throughout its range of motion. Unlocking can be effected in several ways: by rapid cocontraction of the controlling muscles, or "rate" control; by a slower contraction of at least one muscle, or "threshold" control; or by actuation of the same momentary switch that can be used for locking. Lock control by the switch is always available. "Rate" control and "threshold" control of unlocking are mutually exclusive and are determined by an adjustment in the electronics.

When the elbow is unlocked and no myoelectric signals are present, the elbow is in a powered free-swing mode. The free swing is powered (unlike the free-swing modes of the Boston Elbow and the NY-Hosmer Electric Elbow) because the drive transmission of the Utah elbow remains engaged during free swing. Therefore, to overcome the electromechanical inertia of the drive mechanism, the motor actively flexes and extends the elbow, thus drawing battery current, as the arm is swung. The action of the motor is controlled by the response of a load cell transducer to the torque exerted on the forearm.

The Utah elbow can be used in conjunction with other body-powered and electric components having separate control sources. In addition, options exist for integrated control. The electronics added in the Utah Arm version with electric prehensor converts the system from a two-site, two-function controller of elbow flexion and extension to a two-site, four-function sequential controller of elbow and prehensor. In this configuration, the myoelectric sources proportionally control the elbow when it is unlocked. Whenever the elbow is locked, the same myoelectric sources are automatically channeled to proportionally control opening and closing of the prehension device. Unlocking the elbow by "rate" control-rapid cocontraction of the controlling muscles-returns control to the elbow without inadvertent operation of the prehensor.

An electric wrist rotator (Otto Bock) can also be added to the Utah Arm system if there is sufficient forearm length. The rotator can be controlled from a separate and independent source, such as a two-function harness switch actuated by scapular abduction. Alternatively, the movement of the rotator can be controlled by the same myoelectric sources as the elbow and prehensor. In this arrangement, a switch is still needed, but actuation of the switch when the elbow is locked channels the myoelectric signals to the wrist rotator. When the switch is not actuated and the elbow is locked, the myoelectric signals control the prehensor.

Commentary on Electric Elbows

At present, all elbow controllers available through commercial manufacturers control the velocity of motion, either with switches operating the elbow at some preset speed or with proportional controllers (such as myoelectric controllers) that enable the user to directly regulate the speed. Studies by Doubler and Childressindicate that improved control of currently available elbows may be achieved by the use of position servo controllers that directly link movement of a physiologic joint (such as shoulder elevation) to flexion of a prosthetic elbow. In a related study using tracking experiments, it was also shown that position control of a hypothetical prosthesis mechanism with nonlimited dynamic-response characteristics had greater potential for effective control of prosthetic joints than did velocity-control techniques. The implication for the future is that as elbow mechanisms become faster, it will be necessary to adopt different control strategies (than are now used) to take advantage of the improved response. Furthermore, the work suggests that even present-day elbows could be controlled more effectively by position control. These types of controllers have been used experimentally on the NY-Hosmer Electric Elbow and on the Boston Elbow.

None of the available electric elbows approach physiologic performance. There is greater difference between each of the prosthetic elbows and the physiologic elbow than there is among the elbow mechanisms. Therefore, it is not yet possible to truly restore elbow function with these prosthetic components. Consequently, one should consider the many attributes of each of the elbows-including factors such as weight and size, control options, integration into a complete prosthesis, and capacity for being finished in a cosmetically acceptable form-when choosing one for implementation in a fitting.

CONCLUDING REMARKS

The first edition of the Atlas of Limb Prosthetics was published in 1981. Since that time, significant changes have occurred in the types and characteristics of electric-powered upper-limb components. Several devices described in that edition have ceased to be available. Other devices under development then are now commercial items accepted in clinical practice. And still other items have been introduced in the intervening years that were not even conceived a decade ago.

Technological reviews are always in danger of becoming out of date. In several years, the devices and systems described here may no longer be available or may be eclipsed by improved versions. This is the hope. No matter how well accepted or how adequately current devices and systems are thought to perform, they fall far short of the physiologic systems they have been developed to replace.

It is not, however, inevitable that new developments come into being. Many factors beyond technological and conceptual breakthroughs must be brought together to create an environment that supports innovations and provides for the transfer of innovation into clinical practice. How rapidly this section is transformed from a state-of-the-art review to a historical footnote will be a measure not only of the technological advances in our culture but also of the vitality and earnestness of the community working to improve the capabilities of persons who use upper-limb prostheses.

RESOURCE LIST

Hosmer Dorrance Corporation

• 561 Division St
• PO Box 37 Campbell, CA 95008

Hugh Steeper (Roehampton), Ltd

• 237-239 Roehampton Lane
• London, SW14 4LB
• England

Liberty Mutual Research Center

• 71 Frankland Rd
• Hopkinton, MA 01748

Motion Control, Inc

• Division of IOMED, Inc
• 1290 West 2320 South, Suite A
• Salt Lake City, UT 84119

Otto Bock Orthopedic Industry, Inc

• 3000 Xenium Lane North
• Minneapolis, MN 55441

Universal Artificial Limb Co

• 938 Wayne Ave
• Silver Spring, MD 20910

University of New Brunswick

• Prosthetics Research Centre
• Institute of Biomedical Engineering
• 180 Woodbridge St
• Fredericton, New Brunswick E3B 4R3
• Canada

References:

1. Alpenfels EJ: The anthropology and social significance of the human hand. Artif Limbs 1955; 2:4-21.
2. Billock JN: Upper limb prosthetic terminal devices: Hands versus hooks. Clin Prosthet Orthot 1986; 10:57-65.
3. Chan KM, Lee SY, Leung KK, et al: A medical-social study of upper limb amputees in Hong Kong-a preliminary report. Orthot Prosthet 1984; 37:43-48.
4. Childress DS: Artificial hand mechanisms, report 72-Mech-55, in Proceedings of the Mechanisms Conference and International Symposium on Gearing and Transmissions. San Francisco American Society of Mechanical Engineers, 1972, pp 1-11.
5. Childress DS: An approach to powered grasp, in Gavri-lovic M, Wilson AB Jr (eds): Proceedings of the Fourth International Symposium on Advances in External Control of Human Extremities, 1973. Dubrovnik, Yugoslavia, Yugoslav Committee for Electronics and Automation, 1972, pp 159-167.
6. Childress DS, Grahn EC: Development of a powered prehensor, in Proceedings of the 38th Annual Conference on Engineering in Medicine and Biology. Chicago, The Alliance for Engineering in Medicine and Biology, 1985, p. 50.
7. Childress DS, Strysik JS: Controller for a high-performance prehensor, in Proceedings of the 23rd Annual Rocky Mountain Bioengineering Symposium. Columbia, Missouri, Instrument Society of America, 1986, pp 65-67.
8. Datta D, Kingston J, Ronald J: Myoelectric prostheses for below-elbow amputees: The Trent experience. Int Dis-abil Stud 1989; 11:167-170.
9. Doubler JA: An analysis of extended physiological proprioception as a control technique for upper-extremity prostheses (dissertation). Evanston, Ill, Northwestern University, 1982.
10. Doubler JA, Childress DS: An analysis of extended physiological proprioception as a prosthesis control technique. J Rehab Res Dev 1984; 21:5-18.
11. Doubler J A, Childress DS: Design and evaluation of a prosthesis control system based on the concept of extended physiological proprioception. J Rehabil Res Dev 1984; 21:19-31.
12. Engen TJ, Spencer WA: Development of Externally Powered Upper Extremity Orthotics, Final Report. Houston, Texas Institute for Rehabilitation and Research, 1969.
13. Imrhan SN: Trends in finger pinch strength in children, adults, and the elderly. Hum Factors. 1989; 31:689-701.
14. Jacobsen SC, Knutti DF, Johnson RT, et al: Development of the Utah artificial arm. IEEE Trans Biomed Eng 1982; 29:249-269.
15. Jorgensen K, Bankov S: Maximum strength of elbow flexors with pronated and supinated forearm, in Medicine and Sport, vol 6, Biomechanics II, Basel, Switzerland, S Karger AG, 1971.
16. Keller AD, Taylor CL, Zahm V: Studies to Determine the Functional Requirements for Hand and Arm Prostheses. Los Angeles, University of California Department of Engineering, 1947.
17. Kemp MC: Design and development of an electrically powered prosthetic gripper, in Proceedings of the First International Workshop on Robotic Applications in Medical and Health Care. Ottawa, Canada, National Research Council Canada, 1988, pp 19.1-19.5.
18. Komi PV: Relationship between muscle tension, EMG and velocity of contraction under concentric and eccentric work, in Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, vol 1. Basel, Switzerland, S Karger AG, 1973.
19. Mathiowetz V, Kashman N, Volland G, et al: Grip and pinch strength: Normative data for adults. Arch Phys Med Rehahil 1985; 66:69-74.
20. Michael JW: Upper limb powered components and controls: Current concepts. Clin Prosthet Orthot 1986; 10:66-77.
21. Millstein SG, Heger H, Hunter GA: Prosthetic use in adult upper limb amputees: A comparison of the body powered and electrically powered prostheses. Prosthet Orthot Int 1986; 10:27-34.
22. Morrey BF, Askew LJ, An KN, et al: A biomechanical study of normal functional elbow motion. J Bone Joint Surg [Am] 1981; 63:872-877.
23. Northmore-Ball MD, Heger H, Hunter G: The below-el-bow myo-electric prosthesis: A comparison of the Otto Bock myo-electric prosthesis with the hook and functional hand. J Bone Joint Surg [Br] 1980; 62:363-367.
24. Nader, M: The artificial substitution of missing hands with myoelectrical prostheses. Clin Orthop 1990; 258:9-17.
25. Palmer AK, Werner FW, Murphy D, et al: Functional wrist motion: A biomechanical study. J Hand Surg [Am] 1985; 10:39-46.
26. Peizer E, Wright DW, Mason C, et al: Guidelines for standards for externally powered hands. Bull Prosthet Res 1969; 10:118-155.
27. Pertuzon E, Bouisset S: Maximum velocity of movement and maximum velocity of muscle shortening, in Medicine and Sport, vol 6, Biomechanics II. Basel, Switzerland, S Karger AG, 1971.
28. Prout W: The New York Electric Elbow, the New York Prehension Actuator, and the NU-VA Synergetic Prehen-sor, in Atkins DJ, Meier RH (eds): Comprehensive Management of the Upper-Limb Amputee. New York, Springer-Verlag NY Inc, 1989.
29. Ring ND, Welbourn DB: A self-adaptive gripping device: Its design and performance, in The Basic Problems of Prehension, Movement and Control of Artificial Limbs, Proceedings 1968-1969, vol 183. London, The Institution of Mechanical Engineers, 1969.
30. Safaee-Rad R, Shwedyk E, Quanbury AO, et al: Normal

31. functional range of motion of upper limb joints during performance of three feeding activities. Arch Phys Med Rehabil 1990; 71:505-509.

32. Schlesinger G: Der Mechanische Aufbau der kanstlichen Glieder, in Ersatzglieder und Arbeitshilfen, part 2. Berlin, Springer-Verlag, 1919.
33. Sears HH, Andrew JT, Jacobsen SC: Clinical experience with the Utah artificial arm, in The Canadian Association of Prosthetists and Orthotists Yearbook, 1984, pp 30-33.
34. Sears HH, Andrew JT, Jacobsen SC: Experience with the Utah Arm, hand, and terminal device, in Atkins DJ, Meier RH, (eds): Comprehensive Management of the Upper-Limb Amputee. New York, Springer-Verlag, NY, Inc, 1989.
35. Simpson DC: Functional requirements and systems of control for powered prostheses. Biomed Eng 1966; 1:250-256.
36. Simpson DC: The functioning hand, the human advantage, J R Coll Surg Edinb 1976; 21:329-340.
37. Tanenbaum SJ: The Boston Elbow-Health Technology Case Study 29, Report OTA-HCS-29. Washington, DC, U.S. Congress, Office of Technology Assessment, 1984.
38. Taylor CL: Control design and prosthetic adaptations to biceps and pectoral cineplasty, in Klopsteg PE, Wilson PD (eds): Human Limbs and Their Substitutes. New York, McGraw-Hill International Book Co, 1954.
39. Taylor CL: The biomechanics of the normal and of the amputated upper extremity, in Klopsteg PE, Wilson PD (eds): Human Limbs and Their Substitutes. New York, McGraw-Hill International Book Co, 1954.
40. Uellendahl J, Heckathorne CW, Krick H, et al: Four-function hybrid arm prosthesis incorporating an electric wrist rotator and prototype electric prehensor, in Proceedings of the 13th Annual RESNA Conference. Washington, D.C., Resna Press, 1990, pp 167-168.
41. van Lunteren A, van Lunteren-Gerritsen GHM, Stassen HG, et al: A field evaluation of arm prostheses for unilateral amputees. Prosthet Orthot Int 1983; 7:141-151.
42. Westling G, Johansson RS: Factors influencing the force control during precision grip. Exp Brain Res 1984; 53:277-284.
43. Williams TW: The Boston Elbow. SOMA 1986; 1:29-33.

Chapter 6C - Atlas of Limb Prosthetics: Surgical, Prosthetic, and Rehabilitation Principles

O&P Library > Atlas of Limb Prosthetics > Chapter 6C