Prosthesis System for Thumb Amputees (The University of Denver)

The four part assembly of the Thumb Prosthesis System.

Patrick Davenport, Justin Hollenbeck, Joe Tripp


There are a growing number of thumb amputations in the United States. Current prosthetics attach externally to the amputated thumb and are limiting in daily tasks. Intraosseous implants, which are implanted into the bone, may provide a more usable digit with improved feedback to the patient. The purpose of this project is to design an implantable thumb prosthesis that acts as an opposable digit and restores a patient’s ability to perform a variety of daily activities.

The design was divided into three main subsystems: (A) the implant, (B) the implant-prosthesis interface, and (C) the prosthesis itself. The implant was designed to fit a variable population; the interface was designed so the patient can remove the device on demand and quickly; and the prosthesis was designed to allow the patient to perform daily activities and to protect the patient’s bone in the event of an extreme load (e.g. fall).

The system was tested and verified computationally and experimentally to insure that the design met the requirements of the project.


Loss of the thumb results in a 40% impairment of the hand, a 36% impairment of the upper limb, and a 22% impairment of the whole body (1). Today, there exists over 30,000 thumb amputation cases in America alone (2). An improved, thumb prosthesis system is needed to accommodate the growing population of impaired citizens, many of whom are veterans.


Unlike traditional thumb prostheses, an intraosseous transcutaneous (IT) prosthesis system involves an implant that is attached directly to the first metacarpal, protrudes from the digit, and is attached to an external prosthesis. An IT prosthesis replaces the function of the proximal and distal phalanges as a single opposable digit, restoring a patient’s ability to perform a variety of daily activities (Figure 1). The process was pioneered by Dr. Rickard Brånemark and has been utilized with varied success in dental, lower limb, and upper limb applications including the thumb (3).

An x-ray that reveals the bones structure of the human hand.

Figure 1: X-ray revealing the natural thumb bone anatomy.


Current IT thumb prosthesis designs are cumbersome as they often require a tool to attach and detach the external prosthesis, and they facilitate higher rates of transdermal infection. Recent advances in porous metals can promote bone and soft tissue ingrowth, potentially reducing rates of infection.

The purpose of this project is to design the mechanical aspects of a detachable IT thumb prosthesis system that allows patients to safely perform daily activities. While the design was subject to 20 requirements developed with clinical and industry partners, Denver Clinic for Extremities at Risk and OrthoTransmission, the following 5 key requirements imposed the most significant design constraints:

[1] The system shall be able to fit a variable population.
[2] The patient shall be able to attach and detach the external prosthesis in under 30 seconds.
[3] The external prosthesis shall remain attached unless the user intentionally detaches it.
[4] The system shall allow patients to perform daily activities.
[5] The external prosthesis must fail before the implant or the first metacarpal bone when subjugated to extreme loading conditions.

The project was focused on the mechanical design and did not include medical procedures or implant manufacturing. The methods and results of the design are organized into the following three subsystems: the Implant Subsystem, Implant-Prosthesis Interface, and the External Prosthesis (Figure 2).

A simplified diagram of the 3 component thumb prosthesis.

Figure 2: The three thumb prosthesis subsystems.



The main design requirement constraining the implant subsystem (Figure 3) is that [1] the system shall be able to fit a variable population.

A close up of the rendered implant

Figure 3: Implant subsystem


Implant Subsystem Methods

Statistical shape modeling (SSM) was used to determine how the anatomy of the metacarpal changed across a population. Three dimensional representations of the trabecular bone of 35 subjects were segmented from computed tomography (CT) scans using Simpleware (Exeter, UK). The subjects were aligned to a common coordinate frame, and a training set was created by establishing correspondence between the anatomical nodal locations of each subject. Principal component analysis (PCA), a statistical technique that identifies patterns in large data, was then performed on the training set. The SSM identified the key modes and magnitudes of variation of the anatomy of the trabecular bone of the first metacarpal.

From the 35 original subjects, three that best represented the 10th, 50th, and 90th percentile geometries were selected. Implants with varying diameters, lengths, and geometries were overlaid into the three geometries. Implant dimensions were evaluated by the design team and the collaborating orthopedic surgeon, Dr. Ronald Hugate, to determine the implant size range to capture the variability in the population.

Analytical and finite element (FE) methods were performed to determine the minimum acceptable implant radius in the failure region immediately superior to where the implant protrudes from the bone. The methods calculated strain profiles along the implant and bone as a function of implant radius. Loading conditions from Li and Harkness (4) were employed that described a maximum load of 120-N in flexion. The bony material properties were mapped throughout the geometry of the bone based on intensity from the CT scan; the material of the implant was titanium. The bone was fixed at the carpometacarpal joint, and the load was applied in flexion at the most superior point of the implant.

Implant Subsystem Results

Mode 1 characterized scaling and accounted for 46% of the total variability (Figure 4); Mode 2 characterized shaft length and accounted for 33% of the total variability (Figure 5); Mode 3 characterized shaft diameter and accounted for 3% of the total variability (Figure 6). These results indicated that the implant shaft length must extend beyond the isthmus of the metacarpal and that the shaft diameter must be large enough to contact cortical bone.

Head/base scaling of the first metacarpal to two standard deviations.

Figure 4: Mode 1 (46%), head/base scaling.


Length change of the first metacarpal to two standard deviations.

Figure 5: Mode 2 (33%), length of shaft.


Diameter change of the first metacarpal to two standard deviations.

Figure 6: Mode 3 (3%), diameter of shaft.


Shaft diameters of 6, 7, and 8-mm with fluting fill the trabecular region of the metacarpal and contact the cortical bone of the metacarpal (Figure 7). A straight implant shaft of constant length (40-mm) is deemed sufficient for the 10th-90th percentile population (Figure 8).

Three cross-sections of first metacarpals (cortical/implant interaction) for the 10th, 50th, and 90th percentiles of the population.

Figure 7: Three shaft diameters.


Three vertical cross-sections of first metacarpals (cortical/implant interaction) for the 10th, 50th, and 90th percentiles of the population.

Figure 8: Constant shaft length.


A 5-mm titanium diameter yielded at 180-N in flexion indicating that an implant diameter of 6-mm, 7-mm, and 8-mm will not yield under a typical load. Strain in the first metacarpal (2%) was greater than the fracture strain of bone (0.6%). Thus, a failure mechanism is needed to protect the patient from extreme loading conditions.  Similar to hip implants, a coating, e.g. hydroxyapatite, can be applied to the implant to promote bone ingrowth.


The main design requirements constraining the interface subsystem (Figure 9) are that [2] the patient shall be able to attach and detach the external prosthesis in under 30 seconds and [3] the external prosthesis shall remain attached unless the user intentionally detaches it.

The interference interface between the implant head and the prosthesis base.

Figure 9: Implant-prosthesis interface subsystem.


Implant-Prosthesis Interface Subsystem Methods

Two alternative interface mechanisms were considered (Figure 10) before the interference fit mechanism was selected. A computational detachment study was performed using Abaqus (Dassault Systemes, Providence, RI) to determine the necessary interference rib geometry, the force necessary to detach the prosthetic from the implant, and the strain distributions within the prosthesis and the implant head during prosthesis detachment. The implant was fixed in all degrees of freedom, and a displacement loading condition of 11-mm over 5 seconds was applied to the prosthesis (Figure 11). The notch radius and position were optimized to achieve the optimal detachment force.

Two alternative interface mechanisms that the team decided not to use.

Figure 10: Two alternative mechanisms


Implant-Prosthesis Interface Subsystem Results

A notch interference of 1.5-mm radius and 0.1-mm interference yielded a detachment force of approximately 10 pounds (45-N). This was optimal to the design team and the industry advisers.

Figure 11: Detachment Study.

After prototype fabrication, the prototype interface was verified experimentally to meet the requirements of this subsystem. The prosthesis remained attached to the implant unless the user intentionally removed it. If detachment was intended, the prosthesis took less than 3 seconds to be removed from the implant head.


The main design requirements constraining the external prosthesis subsystem (Figure 12)  are that [4] the system shall allow patients to perform daily activities and [5] the external prosthesis must fail before the implant or the first metacarpal bone when subjugated to extreme loading conditions.

The external prosthesis with the failsafe notch identified.

Figure 12: External prosthesis subsystem.


External Prosthesis Subsystem Methods

A literature review was performed to determine measured normal loading conditions experienced by the anatomy of the thumb.

An experimental loading study was performed to understand the extreme loading conditions in the first metacarpal and to verify results from literature. Three subjects of varying weights were asked to perform two actions to subject their thumbs to an extreme loading case. First, the subject was asked to squat down and use their thumb to place as much force as possible on a force plate. Next, the subject was asked to lean back on the ground and get up by pressing their thumb against the force plate (Figure 13).

The author testing the force he can apply with one thumb on a force plate.

Figure 13: Get up motion and resulting force vectors.


A test was performed to optimize the geometry of the failsafe notch mechanism. If an extreme load is experienced, the prosthesis will fracture at the failsafe notch, effectively protecting the bone and implant. Prostheses of varying notch diameters were rapid-prototyped. A transverse load ramp was applied to each until the prosthesis failed at the notch (Figure 14). The failure curve was also calculated analytically and a trend line was applied to the data set.

Experimentally determining the failure points for various diameters of external prosthesis.

Figure 14: Failsafe notch test


External Prosthesis Subsystem Results

Literature from Li and Harkness indicated that the muscles of the thumb can exert a 120-N load when in flexion (4). Wholman and Murray verified Li’s experimental results numerically (5).

Results of the experimental loading study yielded extreme loads ranging from 122.2-N to 287.1-N (Figure 15 and Figure 16). To maintain a safe design, the failure notch must fracture at or below the 122.2-N loading condition.

Graph displaying the force curves for 3 different sized patricipants

Figure 15: Ramp force results


Graph displaying the force curves for 3 different sized patricipants

Figure 16: Get-up force results


The prosthesis will be formed from an injected-molded acrylonitrile butadiene styrene (ABS) plastic material. This material was selected because ABS plastic is affordable, stiff, and can be adapted for testing various prototypes. The prosthesis was designed with a 30° bend to mimic a relaxed anatomical thumb position, making it easier for a patient to perform daily activities. An energy-absorbing silicon sleeve fits over the ABS prosthesis. Fitting of the custom silicon sleeve will be performed by a prosthetic specialist and is the norm for prostheses. The notch-diameter experiment yielded trend lines for printed and injection-molded ABS plastic (Figure 17). The equation for printed ABS (y=0.1513x^3) can be used to predict failure loads (‘y’) by a given notch diameter (‘x’).  Thus, the notch diameter can be customized for an individual patient.

Graph displaying the failure curve as a function of external prosthesis notch radius.

Figure 17: Effects of Notch Diameter on failure load.



After each subsystem was designed a final FE analysis was performed on the entire system to determine failure location, load magnitude at failure, and strain distribution across all three components. The implant was titanium, the prosthesis was ABS plastic, and the material properties of the metacarpal were mapped from CT scans. A 180-N load was applied in flexion at the prosthesis tip and the base of the metacarpal was fixed. At approximately 60-N, the prosthesis exceeded its yield strength, 64-MPa, in the notch (Figure 18). At 120-N and 180-N, the prosthetic experiences approximately 9% and 15% strain at the failure notch respectively. 15% strain is enough to induce fracture within the notch but not in any other location in the system. These results showed that a patient can perform daily activities without risk of breaking their prosthetic. The prosthesis deforms when subjected to a 60-N load and would fracture under a 180-N load. The titanium did not exceed its yield strength (550-MPa) and the strain in the metacarpal did not exceed its fracture strain (0.61%) (Figure 19).

Stress distribution about the failsafe notch on the external prosthesis.

Figure 18: Notch strain.


Strain distribution throughout the bone and implant under given loading conditions.

Figure 19: Bone and implant strain.

The final design (Figure 20) successfully met all of the specified requirements as outlined by the client. The system fits a variable population, is able to be attached and detached quickly and under control, allows the patients to perform daily activities, and protects the patient from extreme cases.

Figure 20: Complete assembly.


Advisors: Ronald Hugate, DC Hoffman, Peter Laz

This research was for, and supported by, Ortho Transmission. The design team would like to thank Dr. Ronald Hugate with Denver Clinic for Extremities at Risk, D.C. Hoffman with Ortho Transmission, and Dr. Peter Laz with the University of Denver for guidance and support. The design team would like to thank the Daniel Felix Ritchie School of Engineering and Computer Science, the Computational Biomechanics Lab, and the Human Dynamics Lab at the University of Denver for facilities and resources.



1. Guides to the Evaluation of Permanent Impairment. American Medical Association. ed 6. 2007.

2. Limb Amputation and Limb Deficiency: Epidemiology and Recent Trends in the United States, Dillingham TR, Pezzin LE, MacKenzie EJ. South Med J. ed 95. 2002.875-83.

3. Osseointegrated Thumb Prostheses: A Concept for Fixation of Digit Prosthetic Devices, Lundborg G, Brånemark PI, Rosen B. J Hand Surg. ed 21A. 1996. 216-221.

4. Circumferential Force Production of the Thumb. Li ZM, Harkness D. Medical Engineering & Physics ed 26. 2004. 663–670.

5. Bridging the gap between cadaveric and in vivo experiments: A biomechanical model evaluating thumb-tip endpoint forces. Wholman SJ, Murray WM. Journal of Biomechanics. ed 46. 2013. 1014-1020.

, , ,

Powered by WordPress. Designed by WooThemes

Skip to toolbar