An Innovative Approach for Shoulder-level Prosthesis Users

Home > Articles > An Innovative Approach for Shoulder-level Prosthesis Users
By Randall Alley, CP

For nearly 30 years I have repeatedly decried the lack of biomechanically based, efficient and stable interface designs and platforms, yet after all this time I still see greater attention to how the interface looks rather than on what the interface can do. While aesthetics, fabrication techniques, and the introduction of 3D printing to O&P is offering improved customization potential, we cannot lose sight of our goal as prosthetists—trying to return to wearers that which was lost. I recently introduced the concept of the Biotensegrity Bridge as a way of optimizing human-device synchronization when designing and assessing interface performance and outcomes. I briefly described the fitting and assessment tools we created and tested as part of our Defense Advanced Research Projects Agency (DARPA) contract, which included a multisite study involving eight subjects: three radial, three humeral, one shoulder disarticulation, and one interscapulothoracic. (Thank you, Summit O&P and Nevada Prosthetics for your participation in this unique endeavor).

In this article, I wanted to highlight our thoracic interface as it required a vastly different approach when compared to humeral and radial tools, designs, and execution. For thoracic interfaces, one has to account for differences in not only range of motion, but also the type and quality of the motion. In limb-based systems, capturing limb motion throughout its range, including changes in direction and speed, requires a different mission set than that of a thoracic based design to effect positioning of the downstream components. While the thoracic design certainly can capture motion (scapular protraction/retraction, elevation/depression, spinal flexion and rotation), it can typically take advantage of a larger footprint with which to stabilize the components throughout their operational motion and functional use. In addition, anatomical joint motion of limb-based systems often contributes a greater percentage of the individual's overall functional envelope than do gross body motions involved in thoracic applications. As such, there is a greater responsibility on the downstream components themselves in thoracic systems to achieve full functional range of motion. Because of this, the primary goal of stability during component operation is often less challenging than when dealing with short lever arms of existing limbs coupled with heavy components.

In the early 90s, I introduced the fully custom X-Frame out of a desire to minimize the total footprint of traditional shoulder-level sockets and as a way of improving user satisfaction and interface performance. I learned the hard way by first attempting what we were taught in school, and quickly observed the massive amounts of interface translation and displacement that occurred with even slight motions of the wearer's shoulder girdle and thorax. I was chasing my tail so much that at one point I simply stopped, stepped back, and asked myself what was the least amount of contact surface I needed to stabilize the system while allowing freedom of movement. The X-Frame was much smaller than existing encapsulated sockets and, as its name suggests, resembled the shape of an X, with its four corners rotated inwardly to form adequate compression of the thorax, and to essentially act as outriggers to stabilize the prosthetic system in any direction. All superfluous material had been removed, including the rigid connection commonly placed over the trapezius. The shape simply provided more with less, with the greatest benefits being improved stability, suspension, comfort, and heat dissipation while also allowing unrestricted motion while concurrently eliminating unwanted interface translation.

The custom X-Frame.

Images courtesy of biodesigns inc.

Given the X-Frame's long-term viability, adoption level in the field, and the success we achieved testing it with physician oversight as part of DARPA's Revolutionizing Prosthetics Program, we made the decision to create a modular version of the X-Frame system, or ModX, for our own DARPA contract. The design could be quickly configured to account for multiple torso sizes and was able to support the weight of the Luke Arm in both 90-degree shoulder flexion and abduction, where forces on the body and the ModX were at a maximum.

The X-Frame design invites the use of a modular, adjustable approach due to it achieving its stability from four stabilizers that allow the rest of the structure to essentially float off the thorax, similar to how a water skipper suspends most of its body off the water, while altogether avoiding the shoulder complex entirely, unless humeral head capture is desired. This allows the use of variable connecting strut segment lengths to position the stabilizers in their optimum location for anthropometric variation.

We strove to emulate the custom design's many benefits while eliminating the challenges a custom-made device presents in a multisite, institutional review board (IRB) study performed in a very short time frame, namely the aforementioned lack of adaptability to different body sizes and the need for extensive fabrication, including casting or scanning, physical or virtual model revision and diagnostic interface creation and assessment.

The ModX prototype.

At first, we considered the ModX to be just a fitting tool, and to be used in such a way that the clinician could quickly size the device to the torso of the subject, mount components, and test stability temporarily.

We quickly realized that using the same leaf joint we use in our floor-standing Imager to prevent inadvertent motion of the paddles under incredibly high forces, we were able to produce a device that could be rapidly fit, easily swappable to either side, and worn with confidence for as long as needed, all while stabilizing a system that was far heavier and more capable than any system that existed at the time.

We designed linkages of different lengths that could be used to reposition the stabilizers and, using our Markforged carbon printer, we 3D printed any part of the system we felt we could. Part of the challenge was that prior to receiving our IRB approval, we had to design the system on a mannequin, which we covered in soft, compressible materials to simulate human tissue, but recognized it was a poor substitute for the real thing. Fortunately, no major redesigns were necessary for the trials and the system worked very well.

The ModX in initial testing.

For testing the ModX, we installed sensors in each of the four stabilizers and, as in the limb-based interfaces, we recorded pressure, rate of rise of pressure (pressure profile), shear, and slippage under various loads in static and dynamic conditions. We documented that recorded shear forces and compression at each of the stabilizers supported the theory behind utilization and acceptance of a modular, distributed, discontinuous surface area approach to thoracic based support of a heavy, multidexterous arm.

The data showed that with extremely heavy loads (up to as much as 13 lb.) and with the subjects performing significant trunk movements with the Luke Arm demonstrator fully extended at the elbow in either shoulder flexion or shoulder abduction, detectable slippage did not occur when using appropriate compression and materials at the stabilizer interfacial boundary and of course a well-designed harness to offset the load. This is extremely important due to most prosthetists still using traditional thoracic shells which encompass the glenohumeral complex (when some remains) rather than avoiding it, resulting in frequent unwanted motion and slippage of the interface, exacerbated with heavier weights and greater shoulder complex motions. (Clinical side note: In cases when a very short humeral application is fitted as a thoracic system, and motion capture of the humeral head is desired, then an XFrame combined with a humeral head extension works well in maximizing volitional motion of the interface while eliminating inadvertent motion.)

The ModX subject testing.

Maximum recorded forces with the shoulder flexed to 90 degrees and all components held in the static condition occurred at the posteroproximal and anterodistal stabilizers. This makes sense given the attachment point at the shoulder and the center of mass of the system out in front such that it induces a rotation moment along the sagittal plane. Our maximum recorded forces with the shoulder abducted to 90 degrees and all components held in the static condition occurred at the anterior and posterior distal stabilizers for similar reasons, with rotational forces occurring along the coronal plane as the center of mass was extended out to the side.

During the trials, we demonstrated we could assemble and fit the test subjects with the ModX in mere minutes, without casting or scanning. The ModX concept proved stable and successful as the joints supported the weight of the Luke Arm in all protocol tests that included multiple weight conditions and body movements, in addition to the subjects stating they were comfortable in the system. While more research is needed, I believe thoracic fittings, be they very short humeral, shoulder disarticulation, or interscapulothoracic in application, should optimize both the footprint and functional contact surface area, and avoid hemispherical containment of the thorax and the shoulder complex if present, which impedes heat dissipation, invites inadvertent interface motion and exacerbates discomfort via unwanted bony contact at the interfacial boundary.

After much of the attention has been placed on components, and in this case, the Luke Arm, over the years, it is my hope that this snapshot of one of the solutions biodesigns developed for prosthetists to achieve what was thought by many to be impossible, or at best extremely difficult—successfully fitting an extremely heavy, multidextrous arm with little upper limb experience—gives you a better appreciation for what went on behind the scenes in our DARPA project. I am extremely proud of all the players involved and although its final effects on the industry remain to be seen, it is my belief that adherence to optimum interface biomechanics trumps all when it comes to ultimate success. It is to this belief I have dedicated my entire career.

To see the DEKA Luke Arm and X-Frame, visit

Randall Alley, CP, is the CEO of biodesigns, Westlake Village, California. He can be contacted at