See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/41825177

Mass Customization of Foot Orthoses for

Rheumatoid Arthritis Using Selective Laser
Sintering

Article in IEEE transactions on bio-medical engineering · March 2010

DOI: 10.1109/TBME.2010.2044178 · Source: PubMed

CITATIONS READS

33 162

3 authors, including:

James Woodburn
Glasgow Caledonian University
137 PUBLICATIONS 1,932 CITATIONS

SEE PROFILE

All in-text references underlined in blue are linked to publications on ResearchGate, Available from: James Woodburn
letting you access and read them immediately. Retrieved on: 24 August 2016
1750 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 57, NO. 7, JULY 2010

Mass Customization of Foot Orthoses for

Rheumatoid Arthritis Using Selective
Laser Sintering
Jari H. P. Pallari, Kenneth W. Dalgarno*, and James Woodburn

Abstract—Rheumatoid arthritis is an inflammatory joint disease Custom orthoses have traditionally been made using craft
that can lead to pain, stiffness, and deformity, often with marked techniques [5], often based on plaster casting an impression
involvement of the small joints of the foot and ankle. Orthotic of the patients’ foot or lower leg, using this “negative” mold
devices are commonly prescribed for this condition to lessen symp-
toms and improve function and mobility, and customized devices to create a “positive,” and then, casting or molding from that
are most effective. The work reported in this paper has examined to create an orthosis. Current state of the art is moving to-
the feasibility of using an additive manufacturing-based approach ward computer-aided design and computer-aided manufactur-
to manufacture customized orthoses. In order to test feasibility, ing (CAD/CAM)-based systems [6] with various centralized
orthoses have been manufactured using the additive manufactur- and distributed models ranging from complete office-based
ing technology of selective laser sintering, and have been evaluated
through a small-scale patient trial (n = 7). The trial indicated that solutions to factory-based manufacturing. These approaches
these orthoses performed as well as the patients’ current prescribed are attempts to move from craft-based customization to mass
customized devices in terms of the observed gait and subjective customization—customization processes, which are systematic
evaluation of fit and comfort. It is concluded that the feasibility and aimed at large markets [7]. However, CAD/CAM is consid-
of the additive manufacturing approach has been demonstrated, ered to raise significant training issues in its application in the
and further development of a mass customization system to deliver
orthoses, together with exploitation of the design freedom offered industry [8], with additional concerns that the design software is
by the manufacturing method, will give the overall approach sig- often basic and the product range is limited by the geometries; it
nificant clinical potential. is possible to create using milling. The aim of this study was to
Index Terms—Additive layer manufacturing, mass customiza- evaluate the potential for an automated and entirely digital mass
tion, orthotics, rheumatoid arthritis (RA), selective laser sintering customization process based on geometry capture using a digital
(SLS). 3-D scanner and manufacture using the additive manufacturing
process of selective laser sintering (SLS). Additive manufac-
I. INTRODUCTION turing processes have been available to produce low volumes
of components with low lead times since the early 1980s [9],
HEUMATOID arthritis (RA) is an inflammatory joint
R disease with an estimated prevalence between 0.3% and
1% [1]. Inflammation causes joint destruction, leading to painful
and have more recently been exploited in the production of
medical devices, including the manufacture of customized in-
the-ear hearing aid shells [10] and the creation of drill guides
and deformed foot joints and orthotics are commonly prescribed for dental surgery [11], and have also been evaluated for their
to redistribute load, restrict or alter motion, or compensate for potential in creating ankle–foot orthoses [12]. One significant
a deformity or muscle weakness. Orthoses have been shown to advantage of additive manufacturing approaches over and above
reduce pain [2], delay the progress of deformity [3] and disabil- the low lead times they operate with is that they offer “design
ity [2], and improve joint function in the foot [4]. Custom-made freedom” [13], with few manufacturing constraints on geome-
orthoses, tailored to the needs of a particular patient, have been try. In order to evaluate the potential for a mass customization
shown to offer improved fit and comfort over mass-produced process based on 3-D scanning and SLS, we have developed a
orthoses [2]. simple rule-based mass customization process. A key element
of any mass customization system is that a customized specifi-
cation can be quickly translated to create a customized product
Manuscript received October 8, 2009; revised November 24, 2009 and design. Typically, this will be based on some modular design
January 25, 2010; accepted February 6, 2010. Date of publication March 4, principles: adding, combining, or adapting different elements of
2010; date of current version June 16, 2010. This work was supported by the a design to create one-off functionality, and the development of
Arthritis Research Campaign under Award W0630 and the European Com-
mission Framework VII Programme under Award CP-PT 228893-2. Asterisk the design rules to support the mass customization approach has
indicates corresponding author. been central to the research. This simple rule-based mass cus-
J. H. P. Pallari is with Materialise NV, Leuven 3001, Belgium (e-mail: tomization process has been then used to create orthoses, which
jari.pallari@materialise.be).
*K. W. Dalgarno is with the School of Mechanical and Systems Engineer- were evaluated by a small sample of patients who have orthoses
ing, Newcastle University, Newcastle upon Tyne, NE1 7RU, U.K. (e-mail: proscribed for them as a result of RA: with the goal of under-
kenny.dalgarno@ncl.ac.uk). standing whether or not these orthoses could match the per-
J. Woodburn is with the School of Health, Glasgow Caledonian University,
Glasgow, G4 0BA, U.K. (e-mail: jim.woodburn@gcal.ac.uk). formance of the patients’ existing orthoses. To our knowledge,
Digital Object Identifier 10.1109/TBME.2010.2044178 this is the first time that a patient trial has been reported with

0018-9294/$26.00 © 2010 IEEE

PALLARI et al.: MASS CUSTOMIZATION OF FOOT ORTHOSES FOR RHEUMATOID ARTHRITIS USING SELECTIVE LASER SINTERING 1751

TABLE I
SUMMARY OF ORTHOTIC DESIGN INSTRUCTIONS INCLUDING SHELL SHAPE AND FEATURES, SPECIAL MODIFICATIONS, AND CORRECTIVE POSTING

TABLE II
SUMMARY OF ORTHOTIC DESIGN INSTRUCTIONS FOR METATARSAL BARS, PADS AND CUSHIONING ELEMENTS

foot orthoses manufactured using an additive manufacturing foot relative to the rest of the foot. Table I summarizes the
process. rules developed for the development of rearfoot wedges
on a particular orthosis.
3) Metatarsal pads, bars, and cut outs. These come in a va-
II. METHODS riety of shapes and sizes and are used to redistribute the
A. Design Rule Development plantar pressure in the forefoot. Table II summarizes the
rules developed for the development of metatarsal sup-
The approach taken in the development of the design rules has
ports on a particular orthosis.
been to use a scan of the patients’ foot to define a “shell,” and
Tables I and II present the outcome of a large and systematic
then, to adapt that basic shell in light of the orthosis prescription.
literature review in order to develop the rules. Full details of the
There were three main features to consider, which are as follows.
process and logic applied can be found in [14].
1) Heel and arch supports. These define a “cup” in which the
heel will be positioned and the arch support, which nor-
B. Mass Customization Process
mally prevents the foot arch from collapsing. Heel sup-
ports are used to deepen posterior support and increase The simple mass customization process used as an exemplar
the effectiveness of arch support, prevent the collapse of in the study is outlined in Fig. 1. Clinical assessment of the
the heel fat pad, and increase the control during the early patient (based on physical examination and gait analysis) was
stages of the gait cycle and reduce pain. Arch supports can used to develop an orthosis specification. In parallel with this, a
also realign the foot and increase the arch contact area in scan of the patients’ foot was taken to give the basic geometry,
order to take pressure from other painful areas of the foot. which an orthosis would be adapted to.
Table I summarizes the rules developed for the develop- Weight and nonweight-bearing scans were taken using a
ment of heel and arch supports on a particular orthosis. hand-held Cobra 3-D scanner (Polhemus, Colchester, VT).
2) Rearfoot wedges. The function of a rearfoot wedge is to Whether to use a weight or nonweight-bearing scan was de-
realign the foot during the gait cycle by rotating the rear- termined by the podiatrist, on the basis of whether the arch
1752 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 57, NO. 7, JULY 2010

The prototype laser-sintered orthoses were benchmarked

against current customized orthotic devices by comparing 1)
the impact on gait and 2) perceived comfort and fit.
The sample comprised three female and four males patients,
with a mean age of 53.4 years (range 29–68 years), mean body
mass of 74.9 kg (range 63.9–96.8 kg), and a mean body mass in-
dex of 26.2 (range 21.4–31.4). The patients have been diagnosed
with RA for 11.3 years (range 1–25 years) on average.
From the patients’ perspective, the trial was completed in two
stages. In the first stage, assessment of each patient underwent a
normal clinical assessment of RA activity in the foot, carried out
by Woodburn following the procedure described by Helliwell
et al. [16], which includes examination of the skin, nail, foot, and
footwear, together with consideration of patient history, tender
and swollen joint counts, and a pain assessment. Patients’ self-
Fig. 1. Mass customization process.
rated pain and orthotic comfort and fit were assessed using 100-
mm visual analog scales (VASs) [4]. This involved the patients
needed to be preserved or supported—in all but one case of making a mark on a simple linear scale, where the ends of the
nonwieght-bearing was chosen. Scans were taken with the pa- scale represented “bad” and “good” fit or comfort depending on
tients seated, with their legs held out horizontally and supported the measure being assessed. A normalized indication of fit or
on a cushioned table, with a Velcro strap holding the leg to be comfort was then obtained by a measurement of where on the
scanned in position on the table. The foot joints were passively line the mark had been made, with 0 designated as “bad” and
manipulated into the subtalar joint neutral alignment while the 100 as “good.” In addition to the normal clinical assessment, the
scans were performed. This pose replicates the standard neutral patients had nonload bearings scans of their feet recorded using
casting techniques used in a high proportion of clinical prac- the Cobra 3-D Scanner with the foot passively manipulated into
tices. Two or three passes of the scanner along the foot were a neutral pose to align the joints.
used to capture the geometry of the foot, with the scanner pro- Once this first stage had been completed, the process out-
prietary software combining data from the passes to produce a lined in Section II-B was followed in order to create an orthosis,
single surface model. which would be fitted to and evaluated by the patient at stage
The foot scan was then exported in .stl format and manip- two. The main aim was a comparison of the orthoses produced
ulated within the Magics CAD software package (Materialise, using the mass customization system in Fig. 1 with the patients’
Leuven, Belgium) in order to meet the specification, manually current orthoses. Laser-sintered orthotics were fitted to the pa-
applying the design rules outlined in Section II-B. Once a de- tient’s own footwear, through placing the orthosis inside the
sign had been completed, the orthosis was produced in Nylon 12 shoe and manually trimming any material from the orthosis,
(Duraform PA, 3D Systems Europe, Hemel Hempstead, U.K.) which prevented it from lying flat within the shoe—in practice,
on a Vanguard SLS machine (3D Systems, Rock Hill, SC). This this meant trimming material from the metatarsal area forward,
combination of material and manufacturing process was chosen depending on the shoe design. The patient’s were requested to
as it allowed for the manufacture of robust orthoses in a material, undergo free walking to evaluate initial fit and to acclimatize to
which is United States Pharmacopeia (USP) class-IV certified the device. This typically took between 10 and 20 min. Patients
(allowing parts to be used in biocompatible and skin contact sit- were then asked to walk (at their normal speed) barefoot, with
uations [15]). After manufacturing, the parts were cleaned and their current orthoses and with the orthoses produced, as shown
finished (typically to smooth surfaces and remove sharp angles in Fig. 1, along a Gaitrite instrumented walkway (CIR Systems,
where that it had not been possible to remove digitally) before Havertown, PA,). This is a 48 × 288 array of pressure sensors
returning to the clinic for fitting to the patient for assessment, encapsulated in a roll-up carpet, which allows the stride length,
which will be next described in Section II-C. walking velocity, cadence (steps/min), and cycle time (time for a
full gait cycle) to be measured. Each patient walked three times
down the Gaitrite carpet for each condition, with the average
C. Patient Trial of the three measurements used for analysis. Patients were then
Seven patients with a definite diagnosis of RA and a current asked to report on comfort and fit of the SLS orthosis using the
history of foot impairments were recruited from the Rheuma- same VAS that had been used to assess their own orthoses.
tology Outpatient Clinic at Chapel Allerton Hospital (Leeds, For statistical analysis of results, two standard tests were used.
U.K.). All patients were in receipt of customized foot orthotic To assess any differences in gait observed between patients, a
treatment on one foot. National Research Ethics Service ap- one-way analysis of variance (ANOVA) test of variance was
proval (Bradford Local Research Ethics Committee (LREC) used. This is relevant, where only one independent variable
Ref: 06/Q1202/36) and Leeds Teaching Hospitals, Research, (variance between groups) is being investigated. In order to
and Development Department approval (Ref: RR06/7554) were compare the results of the comfort and fit assessment, a simple
obtained for the study. paired t-test was used, as this test is appropriate, where the
PALLARI et al.: MASS CUSTOMIZATION OF FOOT ORTHOSES FOR RHEUMATOID ARTHRITIS USING SELECTIVE LASER SINTERING 1753

Fig. 2. Raw foot scan.

Fig. 5. Development of rearfoot wedge. (a) Rotation. (b) Slicing. (c) Final
design.

Fig. 3. Foot scan filtered and processed into a solid model.

Fig. 6. Metatarsal bar design.

Fig. 4. Addition of heel and forefoot extension material to processed foot

solid model prior to trimming. the heel shape. Sharp edges were cutoff so that the orthoses
would fit inside the shoe and not damage it.
Wedging was achieved through creating a series of small steps
same subjects are being tested under two conditions. The level in the arch region in order to provide a graduated transition of
of statistical significance was set at p < 0.05. the required angle, as shown in Fig. 5. To create the rearfoot
wedge—the whole orthosis was rotated to match the wedging
III. RESULTS angle, as shown in Fig. 5(a). Next, the orthosis was “cut” to
create small sections, as shown in Fig. 5(b). The “cuts” were
A. Geometry Capture, Orthosis Design, and Manufacture 1-mm wide and the distance between the “cuts” was 4–5 mm.
Fig. 2 shows a typical “raw” foot surface scan and shows that After these cuts were made, everything else, except the heel
the scan data could be rather noisy. This was filtered and spuri- support part (which was to remain wedged) was rotated 1◦ back
ous data was removed. The scan was then imported into Magics to the original position. Then, the section closest to the heel was
where the surface was distally extruded to create a 5 mm thick kept in place, while the other sections were rotated another 1◦ .
solid model as shown in Fig. 3. The 5-mm wall thickness was This was repeated as many times as necessary to get the forefoot
chosen empirically, as it was found to offer a good combination segment back to its original position. The sections now blended
of strength and stiffness for the application. the wedge to the heel segment while keeping the forefoot flat.
Where necessary material for heel support and forefoot exten- The sections were then joined together to create a single solid,
sion could be added as a block of material, and then, trimmed to as shown in Fig. 5(c).
the required shape (shown in Fig. 4 prior to trimming). In prac- Metatarsal bars were created by elevating “islands” of mate-
tice, only heel supports were utilized in the SLS foot orthoses rial, as shown in Fig. 6, and metatarsal cut outs by removing
as the forefoot extensions, in a solid material construction were material from the model to create recessed volumes. Fig. 7
too stiff to allow necessary toe flexion. The heel support was shows a typical orthosis after manufacture on the SLS machine
trimmed from the top to match the 5-mm wall thickness around and finishing.
1754 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 57, NO. 7, JULY 2010

TABLE IV
ANOVA BETWEEN SC AND SLS FOOT ORTHOSES IN PATIENT TRIAL GAIT
MEASUREMENTS

Fig. 7. Manufactured SLS foot orthosis.

TABLE III
PATIENT TRIAL GAIT MEASUREMENTS (SC, STANDARD CUSTOMIZED FOOT
ORTHOSES; SLS, SELECTIVE LASER SINTERED FOOT ORTHOSES; SD,
STANDARD DEVIATION)

B. Patient Trial
Table III shows the results velocity, cadence, cycle time, and
stride length measured (using the Gaitrite walkway) on the seven
patients for both orthotic treatment conditions. Results showed
only small numerical differences in spatial and temporal gait pa-
Fig. 8. Subjective evaluation of fit of orthoses. Translated directly from
rameters between the orthotic treatment conditions. The mean 100-mm long VAS.
[standard deviation (SD)] walking velocity for the standard cus-
tomized (SC) foot orthoses was 115.1 (19.3) and 113.6 (16.7)
cm/s for the SLS foot orthoses. Cadence was 103.7 (10.2) and for the SC foot orthoses were 79.4 (8.2) mm compared to 77.0
101.7 (9.4) steps per minute for the SC foot orthoses and the (11.1) mm for the SLS foot orthoses. The comfort ratings for the
SLS foot orthoses, respectively. The left and right cycle times SC foot orthoses were 65.7 (19.7) mm compared to 60.6 (21.6)
were 1.16 (0.12) and 1.17 (0.11) s, respectively, for the SC foot mm for the SLS foot orthoses. A simple paired t-test, shown in
orthoses and 1.19 (0.11) and 1.18 (0.11) s, respectively, for the Table V, indicated no statistically significant difference between
SLS foot orthoses. The left and right stride lengths were 133.5 the two sets of results at the p < 0.05 significance level [14].
(17.5) and 134.4 (17.5) m, respectively, for the SC foot orthoses
and 134.7 (14.3) and 134.5 (13.6), respectively, for the SLS foot IV. DISCUSSION
orthoses. Table IV shows the results of a one-way ANOVA test,
which indicates no difference in walking performance of the A. Patient Trial Results
patients using the two orthoses [14]. The main conclusion from the patient trials was that the pro-
Figs. 8 and 9 summarize the subjective responses of the pa- cess had shown that the SLS foot orthoses, manufactured using
tients regarding the fit and comfort of the two sets of orthoses, the process outlined in Fig. 1, had performed as well as the cur-
determined using VAS scales. While there is some variation in rent clinical best practice, etc., while further work is required to
the ratings for a particular patient, the outcome overall in terms develop the concept and approach further, the feasibility of an
of the average response is very similar. The mean (SD) fit ratings additive manufacturing-based mass customization process for
PALLARI et al.: MASS CUSTOMIZATION OF FOOT ORTHOSES FOR RHEUMATOID ARTHRITIS USING SELECTIVE LASER SINTERING 1755

B. Mass Customization Process

The mass customization process used was commercially
naı̈ve. We have made no attempt to develop a valid “supply
chain” as part of this feasibility study. While this has resulted in
a simple system with which to evaluate feasibility, it also means
that quality systems (management systems to ensure product
safety, efficacy, and customer satisfaction) and the flow of prod-
uct data (how digital information and knowledge from scans
and designs would be communicated efficiently and securely
from clinic to orthotic designer to manufacturer) have not been
addressed to the depth required of any medical device.

C. Design Rule Development

Tables I and II present a structured and systematic approach to
Fig. 9. Subjective evaluation of comfort of orthoses. Translated directly from
100-mm long VAS. the development of orthoses, something without which no mass
customization process can function effectively. The initial set
of design rules, based on current clinical practice, showed good
TABLE V
PAIRED t-TEST ANALYSIS OF SUBJECTIVE EVALUATION OF COMFORT AND FIT
face validity, especially for features related to orthotic shape and
functional elements. The rules created close-fitting contoured
surfaces at the heel and midfoot and wedging to meet the thera-
peutic objectives of load distribution, focal pressure relief, and
motion control for feet affected by RA. Further work is required
to refine the stiffness and flexibility features of the sintered ma-
terial relative to the desired requirements on a patient-to-patient
basis. Preliminary statistical analysis supported the observations
of both the clinicians and the patients. We consider the set of
design rules presented as a starting point—further work will
be required to expand and refine the number of rules (in part
through the application of biomechanical modeling) and to test
foot orthoses has been demonstrated. There were two other main them through more comprehensive clinical trials.
findings from the patient trial. One was that the lack of cush-
ioning on the SLS foot orthoses had a significant impact—the
patients perceived the orthoses to be “hard,” and the podiatrist D. Clinical Potential
rejected designs incorporating metatarsal bars, as they were too The adoption of mass-customized orthoses using additive
stiff. It is common with orthoses to apply a layer of cushioning manufacturing approaches will only be widespread if they can
material onto the upper surface, and to have both soft and hard be shown to offer high performance, value for money, and a
areas of the surface. We had hoped that the quality of the fit good service model. The work presented in this paper has shown
would remove the need for these, but that was not the case. It is that orthoses made by SLS to a mass customization model can
possible that the perception of the SLS foot orthoses as “hard” match the performance of SC foot orthoses, but this has not
adversely affected the patient trial results with these orthoses, fully exploited the capability of layer manufacture systems to
but further work would be required to demonstrate this. The sec- locally tailor the structure of an orthosis to provide different
ond point was that the amount of finishing of the orthoses was functional elements. For example, work by Rogers et al. [17]
significant, and therefore, use of software, which would allow has shown within an prosthetic socket design that a device can be
more of the “smoothing” to be carried out virtually, would be of fabricated, which includes low and high stiffness areas, specif-
significant value. ically engineered to give a particular mechanical response, and
It was noted in Section II-B that all but one of the patients this effect could be exploited in foot orthoses in order to mix
were scanned with their foot nonload-bearing. The patient who characteristics from hard and soft orthoses. Other possibilities
was scanned with the foot load-bearing was patient 3. Figs. 8 include the generation of engineered tactile areas on an orthosis
and 9 show that this patient thought the SC orthosis was a better in order to enhance sensitivity in particular areas of the foot (for
fit and more comfortable than the SLS orthosis, and Table III example, the arch regions of the foot are more sensitive than
shows that the patient showed a modest increase in walking the heel, and patients with a high sensitivity in this area show a
speed with the SLS orthosis. Overall, these results are not so good ability to redistribute pressure [18], and therefore, a sur-
different from the others to suggest that the use of a load-bearing face that magnified this effect could have clinical value [19]).
scan had any significant impact on the overall assessment. We Further work will examine the potential for enhanced orthotic
remain of the view that the choice of scan type should be with performance through the use of compliant structures and tactile
the podiatrist. surfaces for function and comfort.
1756 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 57, NO. 7, JULY 2010

The patients who took part in our trial reported paying an [4] J. Woodburn, P. S. Helliwell, and S. Barker, “Changes in three-dimensional
average of £82 (approximately $140) for their existing orthoses joint kinematics supports the continuous use of foot orthoses in the man-
agement of painful deformity in rheumatoid arthritis,” J. Rheumatol.,
(not including clinical costs) [14], and we estimate the design vol. 30, pp. 2356–2364, 2003.
and manufacture costs for producing orthoses using SLS to be [5] M. M. Lusardi and C. C. Neilsen, Orthotics and Prosthetics in Rehabili-
of the order of £50 (∼$80) [20]; therefore, while cost remains a tation. London, U.K.: Butterworth-Heinmann, 2000.
[6] D. G. Smith and E. M. Burgess, “The use of CAD/CAM technology
serious concern within healthcare systems, producing orthoses in prosthetics and orthotics—Current clinical models and a view to the
through this route is not inherently expensive compared to mar- future,” J. Rehabil. Res. Dev., vol. 38, pp. 327–334, 2001.
ket norms. The quality of service, which such a system could [7] R. Duray, P. T. Ward, G. W. Milligan, and W. L. Berry, “Approaches to
mass customization: Configurations and empirical validation,” J. Oper.
provide, would depend on the nature of the commercially de- Manage., vol. 18, pp. 605–625, 2000.
veloped mass customization system. For current additive man- [8] J. P. Otto. (2008, Sep.). A Good CAD-CAM Education Is Hard to Find
ufacturing machines, a central (factory based) manufacturing [Online]. Gainesville, FL: The O&P Edge. Available: http://www.oandp.
com/articles/2008-09_13.asp.
model would be most likely in order to most effectively use [9] G. N. Levy, R. Schindel, and J. P. Kruth, “Rapid manufacturing and rapid
the physical machine capacity, and this would likely make the tooling with layer manufacturing (LM) technologies, state of the art and
minimum lead time for an orthosis 24 h (the orthoses made in future perspectives,” Ann. CIRP—Manuf. Technol., vol. 52, pp. 589–609,
2003.
this study had a one day lead time). However, there is a trend [10] G. Tognola, M. Parazzini, C. Svelto, P. Ravazzani, and F. Grandori, “Three-
toward smaller, cheaper, “desktop” machines, which would be dimensional laser scanning and reconstruction of ear canal impressions for
better suited to a distributed, manufacture at point of prescrip- optimal design of hearing aid shells,” Proc. SPIE, vol. 5009, pp. 19–26,
2003.
tion model [21]. These machines would not be SLS based, but [11] S.-H. Kim, J.-M. Kang, B. Choi, and G. Nelson, “Clinical application of
would employ additive manufacturing principles and have the a stereolithographic surgical guide for simple positioning of orthodontic
potential to bring the lead time down to a few hours. Whether mini-implants,” World J. Orthod., vol. 9, pp. 371–382, 2008.
[12] M. C. Faustini, R. R. Neptune, R. H. Crawford, and S. J. Stanhope,
or not such technologies can deliver the technical performance “Manufacture of passive dynamic ankle–foot orthoses using selective laser
required is yet to be seen. However, experience with CAD/CAM sintering,” IEEE Trans. Biomed. Eng., vol. 55, no. 2, pp. 784–790, Feb.
machines [8] has shown that machines intended for distributed 2008.
[13] J. Giannatsis and V. Dedoussis, “Additive fabrication technologies applied
use must be designed with a nonexpert user in mind. to medicine and health care: A review,” Int. J. Adv. Manuf. Technol.,
Taking the potential for enhanced performance, competitive vol. 40, pp. 116–127, 2009.
pricing, and potential low lead times together, we consider that [14] J. Pallari, “Mass customisation of foot orthoses for rheumatoid arthritis,”
Ph.D. thesis, Dept. Mech. Eng., Univ. Leeds, Leeds, U.K., 2008.
a mass customization system for foot orthoses based on additive [15] (2009, Sep. 28). [Online]. Available: http://www.3dsystems.com/
manufacturing does have significant potential. products/materials/sls/duraformPA.asp
[16] P. Helliwell, J. Woodburn, A. Redmond, D. Turner, and H. Davys, The
Foot and Ankle in Rheumatoid Arthritis—A Comprehensive Guide. New
V. CONCLUSION York: Churchill Livingstone Elsevier, 2007.
[17] W. Rogers, G. Bosker, M. Faustini, G. Walden, R. R. Neptune, and
Overall, we consider that the feasibility study has shown that R. Crawford, “Case report: Variably compliant transtibial prosthetic socket
the prototype mass-customized orthoses produced using addi- fabricated using solid freeform fabrication,” J. Prosthet. Orthot., vol. 20,
tive manufacturing are as effective (within the constraints of the pp. 1–7, 2008.
[18] M. A. Nurse and B. M. Nigg, “Quantifying a relationship between tactile
small-scale patient trial) as currently proscribed orthoses. Fur- and vibration sensitivity of the human foot with plantar pressure distribu-
ther work is required to refine the mass customization system, tions during gait,” Clin. Biomech., vol. 14, pp. 667–672, 1999.
and to establish that an additive manufacturing-based system [19] K. A. Ball and M. J. Afheldt, “Evolution of foot orthotics—Part 2: Re-
search reshapes long-standing theory,” J. Manipulative Physiol. Ther.,
can deliver clinically effective, enhanced performance orthoses. vol. 25, pp. 125–134, 2002.
[20] J. M. Saleh and K. W. Dalgarno, “Cost and benefit analysis of fused
REFERENCES deposition modelling (FDM) technique and selective laser sintering (SLS)
for fabrication of customised foot orthoses,” in Proc. 4th Int. Conf. Adv.
[1] Y. Alamanos, P. V. Voulgari, and A. A. Drosos, “Incidence and preva- Res. Virtual Rapid Manuf., Innovative Dev. Des. Manuf., Leiria, Portugal,
lence of rheumatoid arthritis, based on the 1987 American College of Oct. 6–10, 2009, pp. 187–192, ISBN 978-0-415-87307-9.
Rheumatology criteria: A systematic review,” Semin. Arthritis Rheuma- [21] T. Wohlers, Wohlers Report 2007. Fort Collins, CO: Wohlers Associates,
tism, vol. 36, pp. 182–188, 2006. 2007.
[2] J. Woodburn, P. S. Helliwell, and S. Barker, “A randomised control trial of
foot orthoses in rheumatoid arthritis,” J. Rheumatol., vol. 29, pp. 1377–
1383, 2002.
[3] E. Budiman-Mak, K. J. Conrad, K. E. Roach, J. W. Moore,
Y. Lertratanakul, A. E. Koch, J. L. Skosey, C. Froelich, and
N. Joyce-Clark, “Can foot orthoses prevent hallux valgus deformity in
rheumatoid arthritis? A randomized clinical trial,” J. Clin. Rheumatol.,
vol. 1, pp. 313–322, 1995. Authors’ photographs and biographies not available at the time of publication.