Secondary Logo

Journal Logo


Shoulder Region Socket Considerations

Farnsworth, Troy CP, FAAOP; Uellendahl, Jack CPO; Mikosz, Matthew J. CP; Miller, Laura PhD, CP; Petersen, Branden CP

Author Information
JPO Journal of Prosthetics and Orthotics: July 2008 - Volume 20 - Issue 3 - p 93-106
doi: 10.1097/JPO.0b013e31817d8036
  • Free

When one considers the factors related to successful prosthetic use, most experts would agree that a secure connection between the patient and the prosthesis is a primary factor. Biddiss and Chau1 report on a recent survey of upper limb deficient individuals that describes the respondents’ priorities of factors in prosthetic abandonment. The level of amputation was reported to be the overwhelming primary factor in which persons with high level deficiencies reported 39% and 65% rejection rates for individuals with acquired amputations and congenital deficiencies, respectively. Other critical factors listed were lack of functional benefit, discomfort, too much “hassle”, too heavy, too hot, loss of sensory feedback, and appearance to name a few. Others have reported similar results.2,3 Understanding the factors that contribute to abandonment also can lead to understanding of the factors that contribute to successful prosthetic use.

Because level of deficiency is such a critical factor, careful planning and consideration must be taken to maximize prosthetic use in cases of deficiencies at or near the glenohumeral joint. The prosthetic socket directly relates to the critical factors that have been previously discussed. In addition to traditional encapsulated shoulder design, shoulder frame designs have evolved to improve on these issues. Deficiencies in the glenohumeral region that commonly are treated with variations of shoulder frame design include:

  1. Interscapular thoracic
  2. Shoulder disarticulation
  3. Humeral neck
  4. Very short transhumeral level amputations with comorbid factors
  5. Short transhumeral level amputations secondary to brachial plexus injuries

The use of frame designs for shoulder region prosthetic sockets has been reported for over 35 years and has many contributors to present practice and patient management. Sauter advocated the use of aluminum frames as early as the 1970’s and Ring4 reported on a carbon fiber frame design for fitting shoulder disarticulation prostheses in 1971. The benefits of these designs over sockets that encapsulated the shoulder region include better heat dissipation and lighter weight. By the mid 1980s Andrew advocated the use of a frame design dubbed the “mini frame” for fitting shoulder disarticulation amputees with the Utah Arm using myoelectric control.5 In 1992 Uellendahl and Heckathorne6 reported on a frame design that allowed shoulder motion to serve as an electronic input for hybrid control of electronic shoulder disarticulation prostheses. In 2003, Miguelez and Miguelez7 reported on the use of a frame design dubbed the “Micro-Frame” that minimized motion of the frame, a primary consideration for myoelectric control. Roughly during the same period Alley and coworkers8 advocated a variation of the “Micro-Frame” dubbed “X-Frame”. The use of a frame type socket allows for the requisite stabilization as well as heat dissipation and will minimize weight.

Regardless of the socket design chosen whether an “open” shoulder frame or a “closed” encapsulated design, the goal of the socket must be to provide the most efficient comfortable possible option to “connect” the user to the prosthesis.9 This is best achieved through practical application of the selected design and available materials used for design and construction.

The biomechanics of the shoulder fitting dictate that adequate stability is provided by employing efficient surface area with balanced counterforce to offset the large levers and force couplings created from static and dynamic prosthetic use. In addition to the loading apparent in the sagittal plane, the prosthesis will tend to displace laterally in the coronal plane. The combination of proximal harness counterforce and distal lateral support will counterbalance such forces. The harness system design must act in synergy with the socket design to create the required stability and comfort. Common to all designs a wedge shape proximally resists axial loads. Opposing loads stabilize the socket against rotational forces with force couples in specific regions. Whenever the elbow or shoulder is in a flexed position, a torque is created and is stabilized in the posterior proximal aspect and in the anteriorly distal aspect. Static loading including system weight and functional static loading will apply pressure to the socket and must have adequate counterforce designed into the system. Regardless of the specific frame socket design, the actual dynamic and static loading requirements will be based on the specifics of the system; such as weight, loading during operation, and how it is used (Figure 1A, B).

Figure 1.
Figure 1.:
A, Dynamic loading requirements. B, Static loading requirements.

The type of prosthesis, methods of primary and secondary controls, and user specific requirements will be the greatest determining factors in the specifics of the socket design for glenohumeral region prosthetic fittings.

Primary controls have been defined as terminal device (TD) operation or the movements used to position the terminal device such as the wrist and elbow motions. Secondary controls operate locks, and system controls which typically are essential but performed less often.10 Sequential control is when the same primary control is used to control multiple components. Alternatively each component can have its own dedicated control input. If the system configuration and the user are capable of operating multiple controls together, simultaneous controls and functions may be possible. The socket must be designed to ensure that the interaction of the various controls options and the socket functional requirements are achieved.

If cable-operated (i.e., body-powered) control is employed, the socket should capture as much body motion as possible, particularly biscapular abduction. Any lost motion will limit prosthetic function. Limitations to the motion captured in the system forces the user to overcompensate resulting in increased potential for skin irritation and discomfort. This is further illustrated in case report 1, 5, and 6.

The primary goal when using myoelectric control is to provide a stable platform for acquisition of electromyogram (EMG) signals produced during residual muscle contractions. This will be illustrated in case reports 2, 4, 5, and 6.

Electric components may be controlled through body motions used to activate various inputs. Force sensing resistors (FSRs), switches, force sensors, and potentiometers all are used successfully to control electric components. Motions of a remnant humerus can easily operate various inputs. Likewise ipsilateral shoulder motion has been used to operate FSRs and switches.11 For example, in the case where myoelectric control is being used to operate a terminal device and ipsilateral shoulder motion is activating FSRs to operate a wrist, the socket must serve a more complex dual role. It must provide the stable platform for EMG at the same time allowing the shoulder to move freely and independent of the socket without interference. To further complicate the system, hybrid control may also be employed to operate a cable-operated elbow. In this case, the socket must be able to provide stable EMG signal acquisition, allow shoulder motion within the socket, and capture scapular motion to operate the cable-operated elbow. The case reports illustrate various examples of controls and socket design including case reports 5 and 6.

Various methods and techniques are used to capture the shape of a glenohumeral region socket. Although traditional techniques can be employed, laser surface scanning techniques have greatly improved the efficiency and accuracy of fitting shoulder level sockets. Another crucial step in the shoulder level socket fitting process is the use of a dynamic trial fitting procedure,12 this will allow fine-tuning of the socket parameters and the interaction it has on various controls. This will be discussed in case reports 2, 5, and 6.

With pediatric applications, one must consider typical issues dealing with the growth, developmental needs, simplicity, and the overall functional benefit. Current prosthetic technology offers little functional advantage for the child with a unilateral shoulder level deficiency; as a result the rejection rate is relatively high.13 Important considerations are minimization of weight and ease of control. A simple design that would help the child with basic two-handed grasping activities is a reasonable approach. Based on the experience gained from this first fitting, the team can then proceed with additional functions and active components as appropriate for that individual.14

The bilateral shoulder disarticulation amputee presents several difficult challenges for functional prosthetic fitting. Because of the complexity of such fittings, it is advisable to work with an experienced rehabilitation team to appropriately address all aspects of rehabilitation. Simplicity and reliability should be a priority during the initial prosthetic fitting. If not carefully planned, the combination of primary and secondary controls can be overwhelming. Some consideration should be made to the timing of unilateral versus bilateral fitting and staged introduction of functions and components. It is common to apply varied primary controls to give the user a variety in functions between the prostheses. When the user possesses enough force and excursion for operation of one or more body-powered components, users have demonstrated good functional outcomes when the dominant prosthesis of the bilateral pair is configured with mechanical, cable-actuated components including terminal device and elbow positioning complimented by independently positioned locking wrist motions, humeral rotation, and shoulder joints. The nondominant side incorporates either all electric or hybrid componentry to provide complimentary functions. The electric prosthesis should employ dedicated variable speed control of prehensor, wrist, and elbow whenever possible.15 This is illustrated in case reports 5 and 6.


As technology continues to advance socket designs will also evolve to enable the user to benefit from such advances. Components presently under development are stronger and faster than current systems which will place an increased biomechanical demand on socket systems. Multiple muscle inputs through procedures such as targeted muscle reinnervation and future use of multiple surface sensors for pattern recognition complicate the fitting process. This is illustrated in case report 6. However, new control input methods such as implantable sensors, cortical or EMG, will greatly reduce the complexity of sensor stabilization required using current surface electrodes. Direct skeletal attachment, or osseointegration, may offer another promising development in future prosthetic fitting. Although fittings within the glenohumeral joint region are not practical with the current osseointegration designs, the future may hold expanded options for such an application.



Two socket designs were compared in the effective use of a fully cable-operated shoulder disarticulation prosthesis. The patient was originally fit with an X-frame design where his anatomical shoulder was not enclosed within the socket. This remnant shoulder motion can be beneficial in certain applications where that motion is used to operate a lock or other mechanism within the prosthesis. In situations where an electric locking mechanism is used the remnant motion can be used to apply pressure to a switch or FSR to allow the device to lock and unlock. An electric elbow can also be operated with remnant shoulder motion through the use of a linear potentiometer which could be activated by shoulder elevation. This motion can be performed while still maintaining stability within the socket. These switching or control options only require slight motion from the remnant limb and can be very effective in operating an electric elbow or terminal device. If more excursion is necessary, the X-Frame socket design may not be most suitable. Because the shoulder can move independently within the socket, some of the available excursion is lost. If more excursion is necessary such as in the use of a body-powered elbow which requires approximately 2.5 in of cable excursion, an alternative socket design or a modification to the X-Frame may be more effective in capturing that remnant motion from the shoulder. To achieve this, it may be necessary to enclose the shoulder within the socket to allow the remnant shoulder motion to alter the position of the socket on the limb.16 This repositioning of the socket in relation to the limb can provide the necessary excursion requirements to operate the body-powered elbow. If a full body-powered system is being considered, a thorough evaluation of the patient’s range of motion (ROM) must be assessed. The addition of a body-powered terminal device will require an additional 2 inches of cable excursion which would total 4.5 inches for both the body-powered elbow and terminal device. The ROM and necessary excursion requirements for this type of system may be too difficult for the shoulder disarticulation patient to optimally utilize because of the absence of glenohumeral flexion, which is an excellent source of excursion. This does not mean that a body-powered system cannot be operated by a shoulder disarticulation amputee but it will require some unique harnessing considerations and optimal placement of cable attachments to maximize the excursion potential. Separating the control cables has been done to reduce the excursion requirements for each control cable. Dividing the harness and providing two control cables can be used to generate excursion through scapular abduction and chest expansion. This option proved to be effective in one particular fitting but would most likely not generally be the ideal consideration for control at the shoulder disarticulation level. In one particular fitting, a modified triple control harness was used to capture chest expansion to operate the terminal device and biscapular abduction was used to operate the elbow. This patient was able to operate both the TD and elbow independent from one another. His primary requirement was to position the elbow without reaching over with his sound side, which this setup allowed him to do successfully. He has used a similar type setup for many years so this design worked well for his specific functional requirements. The closed shoulder design was much more functional than the open shoulder design (Figures 2, 3).

Figure 2.
Figure 2.:
Posterior view of cable-operated system.
Figure 3.
Figure 3.:
Lateral view of cable-operated system.


This case study is an illustration of shoulder frame design used to stabilize the electrodes of a sequential myoelectric control system. A surface laser scanning technique produced a cosmetic shoulder fairing.

A 32-year-old man with a nondominant left interscapular thoracic amputation, secondary to sarcoma, presented for prosthetic evaluation and fitting. He is 5 years postamputation, is well-healed, presents with good contralateral ROM, and no comorbid health conditions (Figure 4).

Figure 4.
Figure 4.:
Interscapular thoracic without prosthesis.

A surface laser scanner was used to scan the remnant shoulder to create the model in which the shoulder frame style socket was fit and fabricated (Figure 5). When fitting interscapular thoracic level amputations, careful consideration must be given to provide protection to remnant bony anatomy which is now near the surface of the skin and can be very sensitive. The design principles are similar to standard shoulder disarticulation sockets: to create a stable comfortable “wedge” between the anterior and posterior aspects to transfer loads incurred during prosthetic use. Obtaining stable electrode locations and a comfortable fit were the first priority.

Figure 5.
Figure 5.:
Residual limb scan.

Once the system was proven through a trial fitting phase, a second surface laser scan was taken of his right contralateral shoulder shape (Figure 6). This scan was reflected to produce a mirror image left shoulder shape (Figure 7). From this a cosmetic shoulder fairing was made and applied to the prosthesis17 (Figures 8, 9).

Figure 6.
Figure 6.:
Scan of contralateral shoulder.
Figure 7.
Figure 7.:
Reflected computer model.
Figure 8.
Figure 8.:
IT with finished prosthesis showing cosmetic shoulder cap.
Figure 9.
Figure 9.:
IT with finished prosthesis with clean transition with clothing.


The purpose of this case study is to discuss shoulder disarticulation socket design considerations for an activity specific (AS) cycling prosthesis. There is a gap in the literature concerning this topic because of the small patient population with shoulder disarticulations, and an even smaller portion of that population requiring AS prosthetic devices. Shoulder disarticulation amputations make up only a small percentage of all amputations performed.18

A 45-year-old man presents 23 years postinjury with a left shoulder disarticulation amputation secondary to an industrial accident. His muscle strength and ROM are within normal limits. He has maintained excellent physical health due to competitive cycling activities and reports no health-related issues. He is an accomplished cyclist, competing on a worldwide stage. His occupation consists of teaching in the public school system and teaching competitive racing in his local community. His previous prosthetic experience is comprised of externally powered elbows and terminal devices, he lacks sufficient excursion to operate a body-powered prosthesis due to his short residual limb. He was fit with a new external powered prosthesis, using a force sensor for sequential operation of the system (Figure 10). The socket was designed to capture the scapular abduction for operation of the in-line force sensor (Figure 11).

Figure 10.
Figure 10.:
SD with finished external powered prosthesis.
Figure 11.
Figure 11.:
SD showing in harness force sensor.

In addition to a new externally powered prosthesis, he requested an adaptive cycling prosthesis to minimize sound side arm fatigue while cycling.

In many instances, AS prostheses require repetitive motions for the activity. When designing these types of devices, it is important that the socket is comfortable for the repetitive nature of the specific activity, exhibits excellent suspension, and does not get in the way of the specific ROM required. These AS devices can be very effective for the task for which they are designed. A disadvantage is that the device was designed for a specific activity and may be difficult to use in performing other tasks,19 thus typically requiring two prostheses.

AS sockets can apply atypical force couples from those seen with other prosthetic devices. Close consideration of these forces and how they interact with the user will affect the socket design and alignment of the components. It was discovered in this particular case study that during cycling the socket force couples experienced could be completely opposite form those typically seen and these forces can change quickly. These force couples would change depending where the subject was in his cycling stance. Meaning as this subject initially starts, the socket is pulled down on the residual limb creating force couples typically seen in shoulder disarticulation designs; however, when the cyclist takes a racing stance the socket wants to move superiorly. This created forces couples seen posterior distal and anterior proximal though out the cycling activity. Identifying these force couples and how they interact with the patient is essential to the comfort and utilization of the prosthesis long term.

The following specific design elements were incorporated into this shoulder disarticulation socket. A wedge shaped anterior/posterior compression to minimize distal migration of the socket depicted by (A, B) in Figure 12. The area identified as (C) in Figure 12 approximates the remnant portion of the deltopectoral groove anatomy. This area was accentuated for improved stability. The overall socket footprint was reduced to improve comfort, heat dissipation, and improved cycling ROM.

Figure 12.
Figure 12.:
Reduced size test socket for cycling.

Creating the appropriate amount of socket stability in a shoulder disarticulation socket design is important. In the case of myoelectric socket design typically a more stable socket is desired to support the weight and maintain electrode contact. However in shoulder disarticulation body powered fittings, it could be advantageous to design the socket to be able to be more mobile. Body powered shoulder disarticulation sockets are designed to allow the socket to move with the shoulder to captures more of the required body’s excursion. This subject’s design uses a socket that encapsulates the superior shoulder anatomy for stability purposes as well as captures a small amount of scapular motion. As the subject abducts his left scapula, his residual limb pushes on the inside of the socket and provides extra stabilization as needed. This diagnostic socket design incorporates total contact and total surface bearing socket principles and closely approximates the patient’s residual limb volume and existing anatomy. The rationale for utilizing total contact and total surface bearing concepts was to spreading the forces over his entire residual limb while the subject is stabilizing the socket by scapular abduction. This design criteria proved to be more comfortable over longer durations of cycling motions.

There are many complex, multiaxial motions that occur while cycling. However, the upper extremity ROM experienced by a cyclist with a shoulder disarticulation comes primarily though compensatory, minute transverse trunk (vertebrae) motions to maintain the head and trunk in a forward position, and compensatory shoulder and sound side arm motions to maintain controlled steering of the bicycle. These motions are not significantly noticed during cycling due to trunk and sound side arm compensations. There is also a hip flexion posture when competitive cycling that is maintained throughout the cycling motion. To maintain socket stability for these complex, interconnected cycling motions, the following design considerations were made: smaller socket footprint to be less effected by gross motions, dacron chest strap harness with single pivot harness socket connections to accommodate the sagittal plane motions, surface matched socket contours to improve stability, and a snowboard buckle to allow incremental harness adjustability.

Dacron harness material was chosen over elastic webbing to allow the subject to abduct his left scapula, which pushes his residual limb tighter against the inside surface of the socket without the socket moving laterally away from his residual limb. This provided increased socket stability that the subject could self-adjust by applying pressure.

Figure 13 shows a laminated carbon fiber socket design with views in the anterior, lateral, and posterior views respectively. Figure 13 shows a carbon fiber, laminated socket design in anterior, lateral, and posterior views respectively. Figure 13 also depicts the snowboard buckle, neoaxilla loop padding under sound side arm, and the lower extremity pyramid connection.

Figure 13.
Figure 13.:
Finished socket for cycling prosthesis.

Initial assessment of the socket design and component selection in this case was tested in a controlled manner depicted in Figure 14. The subject assumed the cycling position with the components temporarily attached to assess alignment, prosthesis length, suspension, and socket biomechanics.

Figure 14.
Figure 14.:
Alignment of SD cycling prosthesis.

When designing AS prosthetic devices, it is important to consider safety concerns. Radocy surmises this well. “Safety must be viewed from two perspectives. First, a prosthesis must allow the user to perform activities safely. Second, it must be safe to use around others, protecting them from injury as well.”20

Figure 15 depicts the subject moving the cycling prosthesis into different cycling positions. As the flexible hose bends, more resistance to bending is created, providing a stiffer cycling arm to lean against, and apply forces.

Figure 15.
Figure 15.:
Racing position of SD cycling prosthesis.

The subject is demonstrating the ability to move the cycling arm into different positions to accommodate different racing stances in Figure 15. Because of the properties of this flexible, metal, braided hydraulic hose, as the subject forces the hose to bend, it creates a rigid structure to push against during competitive cycling activities. Figure 16 shows a dynamic assessment of the socket and alignment of the flexible cycling arm. Lower extremity endoskeletal tube clamps and pyramid alignment systems were used for the connections. A Hosmer WE friction wrist was disassembled and the friction elements were mounted to the hydraulic hose connections. A TRS Criterium was used as the terminal device.

Figure 16.
Figure 16.:
Dynamic testing of SD cycling prosthesis.


This case study will be used to demonstrate the use of a shoulder frame design on a prosthesis fit to an individual with a transhumeral level amputation secondary to brachial plexus injury (BPI).

A 23-year-old man presents 4 years postelective dominant right transhumeral amputation. His amputation was performed after a 2 year battle with a flail arm secondary to BPI caused during a motorcycle accident. His glenohumeral joint was fused to reduce pain and stability. No other major comorbid factors effected prosthetic design and fitting (Figure 17).

Figure 17.
Figure 17.:
Transhumeral level BPI without prosthesis.

This patient was previously unsuccessfully fit with a transhumeral hybrid prosthesis; cable operated elbow and switch controlled TD. The patient indicated the rejection was primarily due to the difficulty donning a prosthesis which extended up into his axilla and the he did not have adequate ROM to generate sufficient excursion required for functional prosthetic use.

Given the lack of ROM, an externally powered system was fit. A shoulder frame socket was utilized, allowing simple donning and adequate stability (Figure 18, 19). The patient was fit using myoelectric control, but presented substantial challenges locating and stabilizing the socket over a posterior extensor muscle site. Finally a suitable site was located more distal on his thorax, but the location necessitated a flexible electrode mount to achieve stable electrode contact and sitting comfort. The electrodes were incorporated into the flexible harness system. A friction shoulder joint allowed the prosthetic humeral section to separate from his residual humeral section to allow increased functional ROM and easier clothing donning (Figure 20). This concept can be used with other very short transhumeral level fittings that may not be optimized by a traditional transhumeral design.

Figure 18.
Figure 18.:
Transhumeral level BPI anterior socket frame.
Figure 19.
Figure 19.:
Transhumeral level BPI posterior socket frame.
Figure 20.
Figure 20.:
Transhumeral level BPI finished prosthesis.


Successful prosthetic management of the bilateral shoulder level amputee requires careful evaluation of potential means of controlling the multiple prosthetic components that may be used. Patient evaluation should include locating usable myoelectric sites as well as evaluation of the strength and excursions possible using body motions. Use of these two types of inputs for control is often referred to as hybrid control. When considering the use of myoelectric control in combination with body motions (either for cable-actuated body-powered control or to activate electronic inputs) it is necessary to evaluate whether these control motions can be done independently. Socket design needs to satisfy the particular demands of each control method. For example, if body power is to be used for one or more components, excursion should be maximized by capturing the motion of the shoulder with the socket. If myoelectric control is to be used, the socket should generally avoid capturing shoulder motion so the myoelectrodes will maintain position over the appropriate muscles. When hybrid control is desired a delicate balance between these two socket designs must be achieved.

Because of the complexity of bilateral functional upper limb prostheses, it is advisable to start with a simple system and add complexity as the user becomes proficient in the control and use of the prosthetic components. It is preferable to devise a control scheme that allows the user to control two or more components with control sources that are dedicated to the operation of each component. This type of dedicated control scheme may allow simultaneous and coordinated operation of multiple components providing more natural and spontaneous function.11,15

To illustrate how these principles may be applied to fitting a shoulder level bilateral amputee, the case of MS is presented. MS experienced a high voltage electrical injury. As a result of his injury his dominant right arm was amputated at the humeral neck level and his left arm underwent a shoulder disarticulation amputation (Figure 21). Both lower extremities were amputated at the transfemoral level. For purposes of this discussion only the upper limb prosthetic management will be reviewed.

Figure 21.
Figure 21.:
Showing MS not wearing prostheses.

At the time MS first presented to the amputee rehabilitation team, his right arm was well healed but the left side had not fully healed and presented with extensive scarring over much of the left thoracic area. In consideration of this and to keep the prosthetic system as simple as possible, it was decided to initially fit the right side only. Two muscles suitable for myoelectric control were identified (i.e., pectoralis and infraspinatus). These muscles demonstrated sufficient amplitude for control and could be independently contracted. Movement of the humeral remnant was also evaluated for use as a control input. The humeral movement showed potential for control; however, MS expressed a preference to use myoelectric control finding it easier and less frustrating. Initially the terminal device, wrist rotator, and elbow were controlled sequentially to keep the system as simple as possible. A chin switch selected the component to be controlled. The shoulder joint featured a flexion/extension lock with friction control of abduction. The lock was activated with an electric motor controlled by a chin switch. The scar tissue on the left side was protected with a custom molded, gel lined thermoplastic that prevented the harness from damaging this sensitive tissue by providing a more even pressure distribution.

As MS gained proficiency in use of his right prosthesis, and the scar tissue on the left thorax matured, it was decided to provide bilateral prostheses. To achieve independent control of two or more prosthetic components, MS was reevaluated for useable body motions as well as for myoelectric control sites on the left side. Because the right side was his dominant side, and the right humeral remnant could potentially offer control independent of all other control inputs it was decided to pursue a dedicated control scheme on the right side and use sequential control on the left. Additionally, the available biscapular abduction excursion and force that MS could generate were evaluated. The force was sufficient for body-powered control capable of producing 2 inches of excursion, sufficient to operate a single body powered component. Although myoelectric control had been demonstrated on the right side, continued use of these control inputs was not possible. These muscles either contracted or there was unavoidable electrode motion caused by the body motions used for control with the new design (Figure 22).

Figure 22.
Figure 22.:
Biscapular abduction used for control of the elbow.

Based on the authors’ past experience fitting high-level bilateral arm amputees, it was decided that offering the potential for simultaneous control of the elbow with either the terminal device or the wrist would be most beneficial. On the right side, a control scheme was devised that used biscapular abduction for elbow positioning and humeral neck motion for actuation of force sensitive resistors for control of the terminal device and wrist (Figure 23). Both the elbow and shoulder locks were operated by chin actuated electronic switches. On the nondominant left side myoelectric control operated the terminal device, wrist and elbow in a sequential fashion using chin actuated switch to select the desired component. The locking feature of the shoulder was chin actuated with a switch (Figure 24).

Figure 23.
Figure 23.:
Motion of the humeral neck used to control the electronic wrist and hand with force sensitive resistors. Here a laterally placed FSR can be seen that is activated by glenohumeral abduction. A second FSR is positioned anteriorly and activated by glenohumeral flexion. Each FSR is dedicated to a component using a rate type control scheme to proportionally operate the device.
Figure 24.
Figure 24.:
Finished set of prostheses. The left prosthesis is myoelectrically controlled using a two site system with a chin switch selecting the device to be controlled in a sequential control scheme. The right prosthesis uses a hybrid control scheme that allows simultaneous control of the elbow and either the wrist or the terminal device using motion of the humeral remnant.

After training and much practice, MS became proficient in the operation of his bilateral prostheses. He achieved independence in eating simple foods and in the use of the prostheses to manipulate his environment. The loss of his lower limbs considerably limited his ability to use his arm prostheses because of his inability to position the arms through changes in posture.

The technology utilized in this case, although state-of-the-art, is not capable of replacing the function of the human arm. The authors recognize that achievements made by individuals such as MS are a testament to those individuals determination to succeed coupled with properly designed and fitted prostheses. Clearly provision of optimized prostheses for high-level bilateral arm amputees requires great attention to the details of socket fitting, component selection, and the provision of a reliable control system.


This subject is a bilateral shoulder disarticulation amputee secondary to high-power electrical injury. He was first seen at the Rehabilitation Institute of Chicago approximately 2 months after the injury. He presented with very sensitive skin grafting in the deltopectoral groove bilaterally (Figure 25). There was also healing second-degree burn scarring along the neck and along the sternum. The subject was initially right-hand-dominant and was fit with a body-powered system on this side and an external-powered system on the left. The body-powered system was composed of a Hosmer internal locking elbow, USMC rotation wrist modified to provide dynamic rotation, a Sierra wrist flexion unit and a 5XA terminal device. The elbow and wrist rotation locks were controlled with chin switches. Both sides used LTI/MICA locking shoulder joints; the locks of both shoulders were initially set up with manual chin switches. The externally powered system was composed of a Boston Digital Arm, an Otto Bock wrist rotator and an Otto Bock Greifer. The subject was also initially given hands for both sides, but chose not to use them. The powered side was controlled with 4 FSRs: one mounted superior controlled the elbow, one posterior controlled the wrist rotation and two mounted anterior (one above the other) controlled the terminal device opening and closing.

Figure 25.
Figure 25.:
Photograph of subject approximately 9 months after injury, before surgery to remove painful skin grafts and to perform targeted reinnervation of left pectoralis musculature.

Because of the sensitivity in the deltopectoral groove, it was not possible to fit a socket with loading in this area. The initial sockets were both laminated Sauter frame-style shoulder disarticulation sockets. This perimeter style socket allowed loading through the superior aspect of the shoulder and unloading of the sensitive scarring and the large openings allowed for cooling. The looser fit of this design also allowed movement of the residual shoulder within the socket to contact the FSRs mounted inside.

Because of the painful skin grafts (the subject was even uncomfortable showering due to the sensitivity), a revision was recommended. Because surgery was already part of the clinical plan, the subject was recruited to be the first recipient of targeted muscle reinnervation.21,22 The theory of targeted muscle reinnervation surgery is to create myoelectric sites that match the movement of the amputated arm. The nerves that previously controlled the arm are still present and contain the information about arm movement. Instead of recording directly from the nerve, the nerves are moved to muscles on the residual limb that no longer effect movement (in this case the pectoralis major and minor). These target muscles are deinnervated; the main brachial plexus nerves were moved to the motor points and allowed to connect. Full reinnervation takes approximately 5 to 6 months, after which standard myoelectric electrodes can be used to measure signals and control a prosthesis. The surgery occurred approximately 9 months after the initial injury. During the reinnervation time, the subject was able to continue use of the FSR system.

After reinnervation, it was possible to find three myoelectric signals for control of the prosthesis: elbow up, hand open, and hand close. An elbow down signal was located; however, the magnitude of this signal was smaller than the cardiac interference. Patient strengthening over time and modifications of the electrodes allowed addition of this fourth site in the second myoelectric socket system. All four sites were on the anterior chest wall (Figure 26).

Figure 26.
Figure 26.:
Photograph of left pectoralis, after surgery, showing current electrode configuration.

For the first reinnervation socket, one of the main goals was to create a system that allowed for independent donning. The subject had shown to be a heavy-duty user, though not always using the devices every day. One challenge of the fitting was that during the surgery the pectoralis major and minor were segmented. This resulted in sections of the chest muscle moving independently and in different directions when contracting. Also, since the pectoralis insertion into the humeral head was no longer present, the muscle segments could “ball up” and move up to 1 inch. The spacing of the electrodes for optimal control was less than 1 inch so this degree of movement could cause the control to no longer be functional.

The first socket design was a solid vest. However, even when very tight in the pectoral area, the skin would “pucker” under the contacts. This design was also difficult to don since pushing into the socket would distort the tissue and the electrode locations. A modification of an x-frame or microframe was tried: when stiff and tight enough to hold the sites in place, the socket could not be pushed on comfortably without deforming the tissue. A flexible version was also tried to allow someone to hold the wings open while the subject donned the device and then spring-back onto the chest; however, skin puckering was still present, this degree of pressure was not comfortable, and the design did not allow independent donning.

To make the pressure required for consistent contact more comfortable, a custom silicone pad was eventually found to be the best solution (Figures 27, 28). To fabricate this custom silicone pad, the subject was cast only in the pectoral region. The electrode locations were marked on the cast. This was then filled with plaster and used to create a custom silicone pad approximately 6 mm thick. Dummies were fabricated into the silicone for two LTI packaged electrodes and one Otto Bock electrode. A thin (2 mm) polypropylene layer was also fabricated over the silicone; this rigid layer was limited to the area around the electrodes and where it was designed to connect to the frame. The silicone was connected to the socket using barrel screws and a thin laminate was fabricated to cover the silicone liner to protect the wires. The shoulder area of the socket was left loose to allow movement of the shoulder to contact two FSRs: one located superiorly was used to control an electric lock for the shoulder joint and one located posterior was used to switch control of the two hand electrodes to control of the wrist rotator. The subject used this prosthesis for approximately 2 years.

Figure 27.
Figure 27.:
First version of gel pad, using three electrodes, as shown from the inside. Electrodes controlled hand open and close and elbow.
Figure 28.
Figure 28.:
Second version of gel pad, using four electrodes. A, Check socket version; B, definitive version.

After that time, a decision was made to make a back up socket. By this time, the subject had strengthened the elbow extension signal and modification to the LTI electrodes allowed us to filter the electrocardiogram signal from the elbow extension signal. The introduction of the LTI remote electrodes (preamplifier separate from the skin contact domes) allowed easier fabrication. By this time it was found that the subject did not frequently don the devices independently and this was eliminated as a design requirement. The gel interface had proven to be a success but the subject wanted an easier way to be able to tighten the socket over the area. Therefore, a modified gel pad was developed. With this design, a 3 mm sheet of thermoplastic elastomer (Alpha liner) was cut to cover the electrode area. A thin sheet of polypropylene was again used to connect the gel to the socket, this time by creating a flap. The sheet also prevented the electrode domes from pulling through the gel when tightened. To keep the contacts from sinking into the gel, stainless washers were placed in the hollow backing of the domes. One electrode near the clavicle was fit into the socket. A leather covered plastizote pad was created to protect the wires and wires were routed to the humeral section with a section of conduit cover for the final design. An adjustable strap connected over the pad allowed the subject to adjust the fit of the pad when donning, kept the pad tight over the area of interest, and provided flexibility around the edges for comfort.


1. Biddiss E, Chua T. Upper-limb prosthetics: critical factors in device abandonment. Am J Phys Med Rehabil 2007;86:977–987.
2. Datta D, Selvarajah K, Davey N. Functional outcome of patients with proximal upper limb deficiency—acquired and congenital. Clin Rehabil 2004;18:172–177.
3. Wright TW, Hagen AD, Wood MB. Prosthetic usage in major upper extremity amputations. J Hand Surg [Am] 1995;20:619–622.
4. Ring ND. The Chailey harness with carbon reinforced plastic. Inter-Clin Info Bull 1971;6:5–8.
5. 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, NY: Springer-Verlag; 1989:200–201.
6. Uellendahl JE, Heckathorne CW. Clinical fitting of high level bilateral amputees. Paper presented at the Seventh World Congress of the International Society of Prosthetics and Orthotics, Chicago; 1992.
7. Miguelez J, Miguelez M. The MicroFrame: the next generation of interface design for glenohumeral disarticulation and associated levels of limb deficiency. J Prosthet Orthot 2003;15:66–71.
8. Miguelez JM, Miguelez MD, Alley RD. In: Smith DG, Michael JW, Bowker JH, eds. Atlas of Amputations and Limb Deficiencies. 3rd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2004:263–273.
9. Farnsworth T. Enhancing comfort and function through upper-extremity socket technology. in Motion 2004;14:56–58.
10. Farnsworth T, Alley R. Upper limb externally powered prosthetic designs. In: Carroll K, Edelstein JE, eds. Prosthetics and Patient Management: A Comprehensive Clinical Approach. Thorofare, NJ: Slack; 2006;155–165.
11. Uellendahl, JE. Bilateral upper limb prostheses. In: Smith DG, Michael JW, Bowker JH, eds. Atlas of Amputations and Limb Deficiencies. 3rd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2004:311–325.
12. Farnsworth T. Trial fittings of upper-extremity prostheses. inMotion 2001;11:33–35.
13. Scotland TR, Galway HR. A long-term review of children with congenital and acquired upper limb deficiency. J Bone Joint Surg Br 1983;65:346–349.
14. Uellendahl J, Heelan J. Prosthetic management of children with upper limb deficiencies. Phys Med Rehabil: State Art Rev 2000;14:221–235.
15. Uellendahl JE, Heckathorne C. Creative prosthetic solutions for the person with bilateral upper extremity amputations, In: Atkins D, Meier R, eds. Functional Restoration of Adults and Children with Upper Extremity Amputation. New York: Demos Medical; 2004:225–237.
16. Uellendahl J, Uellendahl E. Body-powered upper limb prosthetic designs. In: Carroll K, Edelstein JE, eds. Prosthetics and Patient Management: A Comprehensive Clinical Approach. Thorofare, NJ: Slack; 2006:141–154.
17. Hanger Orthopedic Group. Insignia Scanner Training Manual, rev 5, 2003.
18. Smith DG. Amputations about the shoulder: surgical management, In: Smith DG, Michael JW, Bowker JH, eds. Atlas of Amputations and Limb Deficiencies. 3rd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2004:251–261.
19. Alley RD, Miguelez JM. Prosthetic rehabilitation of glenohumeral level deficiencies. In: Meier RH, Atkins DJ, eds. Functional Restoration of Adults and Children with Upper Extremity Amputations. New York: Demos Medical; 2004.
20. Radocy R. Prosthetics adaptations in competitive sports and recreation. In: Smith DG, Michael JW, Bowker JH, eds. Atlas of Amputations and Limb Deficiencies. 3rd ed. Rosemont, IL: American Academy of Orthopaedic Surgeons; 2004:327–337.
21. Kuiken TA, Dumanian GA, Lipschutz RD, et al. The use of targeted muscle reinnervation for improved myoelectric prosthesis control in a bilateral shoulder disarticulation amputee. Prosthet Orthot Int 2004;28:245–253.
22. Lipschutz RD, Kuiken TA, Miller LA, et al. Shoulder disarticulation externally powered prosthetic fitting following targeted muscle reinnervation for improved myoelectric control. J Prosthet Orthot 2006;18:28–34.

shoulder; shoulder disarticulation; forequarter; interscapular thoracic; humeral neck; interface design; socket design; shoulder prosthesis; brachial plexus injury; bilateral upper limb; targeted muscle reinnervation

© 2008 American Academy of Orthotists & Prosthetists