Christopher Welch, Casey Beasley, Joseph Cassella, Frank Finelli, Stephen Sousa, Michelle Swanston
ABSTRACT
Today’s prosthetic hands incorporate stock digits in a limited number of sizes. But finger size has implications for aesthetics, hand functionality, and rehabilitation. Is there a need for a prosthetic phalanx that can dynamically adjust to the patient-specific size and shape preferences? Furthermore: where no such finger has ever been brought to market, is a customizable prosthetic phalanx (CPP) even possible? We believe the answer to both questions is: “YES!”
We have designed a prosthetic finger that is fully adjustable in terms of both circumference and axial dimensions, while retaining the full function expected of an off-the-shelf digit. Here, we describe our design and development of a working prototype; in addition to photo and video demonstration of the prototype, we provide a full business plan including market analysis and future outlook, and a draft manuscript describing the anthropomorphic implications of right-sizing a prosthetic hand. We believe that a CPP would transform the lives of underserved populations in the upper-limb prosthetics market, i.e. women, children, users with extremely technical or dexterous design needs, or users seeking a restoration of their limb identity. In this way, we believe that the CPP will reduce device abandonment, delivering new technological promise to those patients who are simply dissatisfied with the look and/or performance of the currently available options. With the ability to reshape the CPP into sizes big, small, fat, slim, long, and short, we believe that patients with upper-limb deficiency will see the CPP as “More Like Me.”
The Customizable Prosthetic Phalanx
INTRODUCTION
Problem Statement: Device abandonment rates for upper-limb prostheses vary widely in the peer-reviewed literature, but are commonly reported at 25-50% [doi: 10.1097/PHM.0b013e3181587f6c, 10.1016/j.apmr.2011.11.010, 10.1080/03093640600994581]. Critical design factors which would increase device utilizations are cosmetic appeal, functionality, and incorporation into the patient identity [10.1097/PHM.0b013e3181587f6c, 10.1080/17483100701714733, 10.1080/17483100601138959].
Motivation: Currently available prosthetic hands incorporate digits of stock sizes that are representative of the 50th percentile male (Todd Kuiken TED Talk @ 17m15s), and even then: not necessarily so. The human hand is highly variable, as seen across (and within!) major anthropometric studies, e.g. in the Metacarpal phalangeal joint to finger tip distance (MCP to tip) in males:
MCP to Tip (Middle Finger, anthropometric studies, mm units): 83.8 +/- 5.4 [10.4103/0970-9290.117993], 81.3+/-7.1 [DTIC: ADA244533],69.8+/-4.31 [10.1016/j.ergon.2008.01.010].
(More comprehensive review of human size measurements is available in our draft manuscript.)
Whereas popular prosthetic hands yield very different values: MCP to Tip (Middle Finger, prosthetic hands, mm units): 87 (iLimb Ultra), 103 (BeBionic3), Shadow Dextrous Hand (96mm), and 97mm (IH2 Azzurra).
Some manufacturers (iLimb) can populate a single hand with four different digits:
iLimb allows up to four different finger sizes.
While this is very good, these are four discrete sizes, akin to Small-Medium-Large-Extra Large. We believe patients would prefer to have custom sizes. We note that other hands, e.g. Prensilia, use the same digit at all four finger positions
Prensilia uses the same digit at all four positions.
The human is simply too variable and too personal; the market-available hands cannot possibly reflect user needs on a per-patient basis.
DESIGN AND DEVELOPMENT
Early Designs
Our first instinct was a cylindrical clamp-style phalangeal segment for length adjustment; width adjustments would be allowed passively by stuffing the clamp with spacer.
Design 1a
Design 1b
Design 1c, with clamp-style phalangeal segments for length adjustments; Spacer for width adjustment not shown.
We decided that this was not a sleek design (early prototypes with plumbing clamps were very rough visually and tactilly). We then created a series of intermediate prototypes that did not satisfy our needs (for brevity, images replace discussion).
Intermediate Design A
Intermediate Design B
Intermediate Design C
Intermediate Design D
We then decided on a core-and-wing system, which allowed for simultaneous adjustment axially and radially, with a sleeker profile.
Design 2, with radially affixed wings for diameter expansion.
Design 2
This would serve as the basis for the most current design, which modifies the inner core slightly, to provide a stability bar for greater structural integrity in the interlocking components.
Design 3a, with stability bar (t-tab on gold-colored cylinder at left in both panels).
Design 3b
These designs are for phalangeal segments; they will be connected by simple hinge joints (design not shown).
Design 4a: Full finger with two adjustable barrel units (light blue and white) and a non-adjustable finger tip (orange). This is the full design of a CPP (set screws not shown).
We note that in order to assemble a full hand, we had to integrate the CPP into the existing Hartford Hand platform, with minor alterations to the platform. We highlight a few of the collateral design modifications here.
Design 4c: Collateral design modifications to the Hartford Hand in support of the customizable prosthetic phalanx (CPP) design. Shown in Design 4: Palmar aspect (front view), Palmar aspect (rear view), Metacarpal phalangeal joint tray, and thenar element.
Design 4d
Design 4e
Design 4f
Design 4g
Physical Prototypes
In the early stage development of working prototypes, we attempted 10 different physical prototypes.
Prototype 1: Early prototypes with plastic tubing.
Our first prototype to incorporate a palmar unit was then assembled using fingers comprising plastic and wood.
Prototype 2a
Prototype 2b
Prototype 2c: Interim prototypes with wood and plastic tubing, based on one of the interim finger designs (Interim Design A).
The current working prototype reflects a 3-D printed version of the latest design.
Side view of interlocking pieces
Top view of interlocking pieces
Fingers joined on dorsal side of hand.
Fingers joined on dorsal side of hand (wiring is for flexion/extension movement).
Design 3b, with stability bar (t-tab on gold-colored cylinder at left in both panels). Resembles basic structure of current prototype.
We note that the finger was 3-D printed to approximately 150% scale so that the prototype could be easily trouble-shooted. Our plan is to scale to 100% anticipated unit size in the next prototyping. One limitation here is that the commercial 3D printing service we contracted with provides affordable prototyping, but does not provide high-resolution printing. For this reason, all parts are thicker than we would like; in this way, it is possible to obtain a finger fit for a full grown adult in this version, but not for a child. In future prototypes, we anticipate that a higher-precision printing, or machining will allow thinner pieces to be assembled, allowing us to service even the smallest hands. Nevertheless, the finger is adjustable in both dimensions, as desired. The prototype phalanges are now incorporated into the Hartford Hand, where their full flexion ability is demonstrated. Please refer to the video for a visual demonstration of the adjustability in length, circumference and the microprocessor used to provide flexion and extension of the finger.
The following video is an animated demonstration on the Hartford Hand.
DISCUSSION
Competing Technologies
In order to guide our product development, we created a partnership with a student in our University’s Business School with an interest in understanding the market forces that would steer our product’s development, viability, and ultimately: delivery to market. From this market analysis, we identified several technologies of interest, including a patented “enhanced functional prosthetic limb” (USPTO: US20050234564). This technology in particular is not a threat to our device, as it describes axial elongation by a different mechanism, and does not describe radial expansion. Furthermore, we have surveyed the market for all available hands, and obtained hand/finger measurements from as many manufacturers as possible. Comprehensive company and product descriptions are available in this business plan.
Market Considerations
The forecasted compounded annual growth rate for the orthotics and prosthetics global market is 6 percent for 2010 through 2017, which will reach approximately $4.5 billion by the year 2017. While we acknowledge that this is a combined market of orthotics and prosthetics, and furthermore that prosthetics could include upper-limb or lower-limb, market figures for the upper-limb prosthetics market are not well-established. However, it is estimated that market share for prosthetics (versus orthotics) is approximately 26%.
There are an estimated 26,000 upper limb amputations each year. Hands typically sell for $10,000-$30,000, but the Shadow Robot Company’s Dextrous Hand sells for $120,000. An entire Hartford Hand (complete with CPPs) can be manufactured for an estimated approximately $200 per unit; we anticipate a viable market price of approximately $10,000. Furthermore, we assert that the fingers themselves can be made to retrofit onto existing hands, and could easily be designed as accessory parts for any hand manufacturer. Comprehensive pricing and market review information are available in our business plan.
Patient Impact
Is precision sizing necessary? We believe so. Ramachandran’s Mirror Box [doi: 10.1093/brain/awp135] shows that biomimetic visual feedback greatly reduces phantom limb pain, and reduces potential problems due to “dimensional mismatch” (as might occur if the prosthetic hand were improperly sized) [10.1068/p7569, 10.1016/j.concog.2003.07.001]. Hands that look only somewhat like the patient’s true hand run the risk of stepping into the ‘uncanny valley’, i.e. the steep decrease in user satisfaction as objects become nearly –but-not-quite- life-like [10.1109/MRA.2012.2192811]. Please refer to this link for more information.
Uncanny Valley for prosthetic hands shows that nearly-life-like hands are the most “eery”, while hands that look truly like human hands are the most acceptable. Source: Poliakoff, et a., 2013, doi: 10.1068/p7569.
We believe that patient identity is paramount in reversing the trend of device rejection. Upper-limb prostheses have a deeply personal meaning to the user, and have strong implications for personal sense of identity [ISBN: 0387874623, doi: 10.3109/09638288.2012.723790]. In this way, we believe that the psychological impact of a natural, life-like hand would be similar to the value of proper breast reconstruction or prosthesis after mastectomy; there is extensive evidence that patients are keen to “look like their old selves” [10.1080/07399339409516129, 10.1097/01.prs.0000278162.82906.81]. Considering that the hand is not only externally viewable, but is the only vehicle by which the patient will be able to manipulate objects, shake hands, and write on paperwork, should not the prosthetic hand be equally availed to customized sizing? For more discussion on the patient impact of our design, we refer to our working manuscript.
Acknowledgements
We would like to thank the Women’s Education Leadership Fund at the University of Hartford, the Connecticut Space Grant Consortium, University of Hartford Prosthetics & Orthotics Program, Department of Rehabilitative Sciences, College of Education, Dr. Barbara Crane @ RESNA, Stephanie Hebert, Derek Becker and Chelsea Dornfeld. Special thanks to faculty mentor Dr. Michael Wininger for supporting this project. Lastly, we would like to thank RESNA for receiving and reviewing our submission.
About the Team
Our team comprises students in the University of Hartford Master’s of Science in Prosthetics Program. Casey Beasley, Joseph Cassella, and Michelle Swanston are in the program’s Year 2; Stephen Sousa and Christopher Welch are in Year 1. Frank Finelli is a senior in the Barney School of Business at the University of Hartford. This project was completed via three sub-groups: a mechanical design team (CB + SS), a hardware team (JC + CW), and a business plan team (MS + FF); the goal was to pair a graduating Master’s student with a younger student who will be able to grow the project after the Year 2 students leave for their clinical residencies.
Contact Information:
Christopher Welch
christopher.welch@outlook.com