Evaluation of prominence of straight plates and precontoured clavicle plates using automated plate-to-bone fitting

Abstract

Hardware prominence after plate fixation for clavicle fracture is a common complication. The aim of the study was to perform a 3D analysis of the prominence of different types of superior clavicle plates. An automated fitting of 3 straight and 10 precontoured plates was performed on 52 3D-CT-scan reconstructed cadaver clavicles. The mean and maximum bone-plate distance and maximum prominence was significant higher with the straight plates compared to the precontoured plates. The mean and maximum boneplate distance was significant higher with the precontoured DePuy-Synthes plates compared to the precontoured Acumed plates but when evaluating the maximum prominence there was no significant difference between the most commonly used 8-holes plates. To conclude, precontoured plates of the clavicula diminish significantly hardware prominence. There exists a difference in hardware prominence between different brands of precontoured plates but this difference is limited and in most cases not significant.

Full text

Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
Hardware prominence after plate xation for clavicle
fracture is a common complication. The aim of the
study was to perform a 3D analysis of the prominence
of different types of superior clavicle plates. An auto-
mated tting of 3 straight and 10 precontoured plates
was performed on 52 3D-CT-scan reconstructed ca-
daver clavicles. The mean and maximum bone-plate
distance and maximum prominence was signicant
higher with the straight plates compared to the pre-
contoured plates. The mean and maximum bone-
plate distance was signicant higher with the precon-
toured DePuy-Synthes plates compared to the
precontoured Acumed plates but when evaluating the
maximum prominence there was no signicant differ-
ence between the most commonly used 8-holes plates.
To conclude, precontoured plates of the clavicula di-
minish signicantly hardware prominence. There ex-
ists a difference in hardware prominence between dif-
ferent brands of precontoured plates but this
difference is limited and in most cases not signicant.
Keywords : clavicle fracture ; hardware irritation ; pre-
contoured plates ; 3D reconstruction ; automated fitting.
INTRODUCTION
Approximately 2% to 5% of all fractures in adults
involve the clavicle. More than two-thirds of these
injuries occur at the diaphysis of the clavicle, and
these injuries are more likely to be displaced as
compared with medial- and lateral third frac-
tures (17,18). Recent evidence suggests these specic
subsets of patients may be at high risk for nonunion,
shoulder dysfunction, or residual pain after non-
surgical management (6,14,24,30). In these patients,
acute surgical intervention may minimize subopti-
mal outcomes. Internal xation of clavicle fractures
can be performed with either plate or intramedullary
pin xation with good result (22). Most complica-
tions in both groups are hardware-related. Irritation
with subsequent removal of the hardware is the
most common cause of additional surgery (28,29).
Where intra-medullary devices are routinely re-
moved upon fracture healing, the plates are also
No benefits or funds were received in support of this study.
Conict of interest: We want to thank the BVOT (Belgische
Vereniging van Orthopedie en Traumatologie) for their nan-
cial support and Iwein Piepers for his statistical support.
Acta Orthop. Belg., 2014, 80, 301-308
Evaluation of prominence of straight plates and precontoured clavicle plates
using automated plate-to-bone alignment
Alexander Van Tongel, Toon Huysmans, Bernat amiT, Jan sijbers, Francis Vanglabbeek, lieven De WilDe
From the University Hospital Ghent, Ghent, Belgium
ORIGINAL STUDY
n Alexander Van Tongel, MD, PhD, (Orthopaedic Surgeon).
n Lieven De Wilde, MD, PhD (Orthopaedic Surgeon).
Department of Orthopaedic Surgery and Traumatology,
Ghent University Hospital, Gent, Belgium.
n Toon Huysmans, PhD (Engineer).
n Bernat Amit, MD (Orthopaedic Surgeon).
n Jan Sijbers, PhD (Engineer).
Vision Lab Department of physics, Antwerp University,
Belgium.
n Francis Vanglabbeek, MD, PhD (Orthopaedic Surgeon).
Department of Orthopedic Surgery and Traumatology,
Antwerp University Hospital, Belgium.
Correspondence : Alexander Van Tongel, Department of
Ortho paedic Surgery and Traumatology, Ghent University
Hospital, De Pintelaan 185, 9000 Gent, Belgium.
E-mail : Alexander.vantongel@uzgent.be
© 2014, Acta Orthopædica Belgica.
van tongel-.indd 301 29/09/14 10:43

302 a. van tongel, t. huysmans, b. amit, j. sijbers, f. vanglabbeek, l. de wilde
Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
frequently removed due to irritation. A second
operation with plate debridement, removal or revi-
sion is required at best in one out of every ten
patients treated, and in some studies even up to one
out of every two patients (2,5,6,11,12,20,21,23,25).
To address this complication precontoured clavi-
cle plates were introduced. The low prole and
beveled edges of the plates were thought to have a
better plate-bone contact resulting in a reduction of
the incidence of irritating hardware prominence and
the need for reoperation for hardware removal. The
aim of this study is to determine the prominence
of commercially available, straight versus pre-
contoured superior claviclar plates, using a three
dimensional 3D-CT scan reconstruction analysis
and an automated plate-to-bone alignment.
METHODS
Fracture Fixation Plates Sets
First 3 straight plates (S) (6 – 7 and 8 holes) were cre-
ated. The length, width and thickness of these plates were
based on LCP Plates (DePuy-Synthes). The thickness of
the plate was equivalent all over the plate (3.3 mm). The
location of the holes were equivalent distributed over
these custom plates (Fig. 1). Next 3 companies with pre-
contoured plates (Acumed, DePuy-Synthes, Smith and
Nephew) were contacted. Both Acumed and DePuy-Syn-
thes provided us with accurate descriptions (STL les) of
the three dimensional geometry of their plate sets for
superior xation of mid-clavicle fractures (Fig. 1). Note
that, contrary to the Acumed plates which are only curved
in the axial plane, the DePuy-Synthes plates are also (lat-
erally) curved in the frontal plane. For each of these
plates the plate-to-bone contact region and the position of
all the screw holes were extracted automatically from the
STL-le. Note also that concerning the thickness of the
plate, in contrast to the straight plates, the thickness in
precontoured plates is different at the side compared to
the middle.
Study Population
In this study, 52 clavicles from 52 distinct human
Caucasian cadavers were dissected. This set of clavicles
represented 32 male and 20 female specimens with a
mean age of 71 years (range : 25 to 99 years). The popu-
lation consisted of 50 (31 male, 19 female) right and 2
(1 male, 1 female) left clavicles.
Data Acquisition and Preparation
Preparation of the clavicles was done in the anatomy
lab of the University of Antwerp. All clavicles were
scanned with a GE LightSpeed VCT (GE Medical Sys-
tems, Milwaukee WI, USA) with a spatial resolution of
500 × 500 × 600 µm3 at the Antwerp University Hospital.
The computed tomography (CT) reconstructions from
the GE Lightspeed Volume CT system were automati-
cally segmented by morphological image-processing
operations. From the obtained segmented images, the
outer boundary surface of each clavicle was extracted
using the marching cubes algorithm (13).
Finally, all right clavicles were mirrored with respect
to the sagittal plane and thereby brought into the coordi-
nate space of the left clavicle.
Common Reference Coordinate System
In order to facilitate an automated plate-to-bone tting
procedure, all clavicles are placed in a common reference
coordinate system following the three-steps procedure of
Huysmans et al (9,10).
Automated Plate-to-Bone Fitting
The common reference coordinate system enables the
automation of tting a given plate to a given clavicle with
optimal plate-to-bone contact while satisfying several
constraints imposed by the surgical procedure. On the av-
erage clavicle a single point was placed manually on the
superior part in the middle of the clavicular surface as
dened in the reference system. A fracture was simulated
by cutting the clavicle along the angular line of the cylin-
drical coordinate system that runs through the annotated
point. This is followed by the calculation of the desired
region of contact for the plate. This is a region of 100mm
in length and 10mm in width dened along the axial line
that runs through the annotated point (Fig. 2). The frac-
ture line and the desired region of contact, as dened on
the average clavicle, can be mapped using the correspon-
dence to each of the 52 individual clavicles in the popula-
tion, effectively simulating 52 fractured clavicles.
For a given clavicle, the automated plate tting proce-
dure proceeds in two steps. First, an initial alignment of
the plate to the bone is obtained by aligning the center
and principal axes of the contact region of the plate to the
center and principal axes of the desired contact region of
the bone, taking into account the medial and lateral sides
of the plate. After this initialization, the plate and bone
may intersect and other constraints, imposed by the
surgical procedure, may not be satised. We therefore
van tongel-.indd 302 2/10/14 10:44

Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
eValuaTion of prominence of sTraigHT plaTes anD precounTereD claVicle plaTes 303
introduced a second step. This step is an optimization
that minimizes the distance between plate and bone,
measured as the mean of squared distances from the con-
tact region of the plate to the closest points on the bone.
During this optimization the following constraints are
also enforced : (a) avoid intersection of plate and bone,
(b) ensure at least three screws on each side of the frac-
ture with a minimum distance of 4mm from the fracture
line, and (c) ensure that each screw catches enough bone,
i.e. the centerline of the screw should be at least 3mm
from the side of the bone. When the optimization does
not succeed in satisfying all the constraints, the plate is
considered a bad t for that specic clavicle (Fig. 3).
The number of good and bad ts for every plate was
measured. Next the plates were grouped in group A
(6 holes), group B (7 holes), group C (8 holes). The plate
of Acumed with 10 holes was excluded because no com-
parison could be made with an equivalent plate of Depuy-
Synthes or straight plate. We did not grouped the plates
concerning their length because during surgery the sur-
geon seems to be more guided by the number of holes
then by the length of the plate. Next the mean plate-to
bone distance for the different plates on every clavicle
was measured. The next step was to measure the maxi-
mum bone-plate distance for the different plates on every
clavicle. These two distances are a measure of how tight-
ly the plate ts to the bone (Fig. 4). At last the maximum
hardware prominence for the different plates was mea-
sured as well. This was measured as the largest minimum
distance between the plate and the bone. This measure-
ment gives an estimate of the largest tissue displacement
due to the plate (Fig. 4). The statistical analysis was per-
formed using Chi² test, the Fisher’s Exact test, Mann-
Whitney U test and the Kruskal Wallis test.
Fig. 1. — 13 different tested clavicula plates
Fig. 2. — Calculation of desired region of contact for the plate
van tongel-.indd 303 29/09/14 10:43

304 a. Van Tongel, T. Huysmans, b. amiT, j. sijbers, f. Vanglabbeek, l. De WilDe
Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
The mean maximum distance of every plate can
be seen in Table I. The mean maximum distance is
signicant higher with the straight plates compared
to the precontoured plates in group A (p < 0.001), B
(p < 0.001) and C (p < 0.001). The mean maximum
bone-plate distance is signicant higher with the
precontoured Depuy-Synthes plates compared to the
precontoured acumed plates in group A (p < 0.001),
and C (p < 0.001).
The mean maximum hardware prominence of ev-
ery plate can been seen in Table I and gure 5. The
mean prominence is signicant higher with the
straight plates compared to the precontoured plates
in group A (p < 0.001), B (p < 0.001) and C
(p < 0.001). The mean maximum hardware promi-
nence is signicant higher with the precontoured
DePuy-Synthes plates compared to the precon-
toured Acumed plates in group A (p < 0.001), but
not in group C (p = 0.054).
DISCUSSION
This study determines the prominence and its
maximal location of two commercially available
precontoured superior claviclar plates (DePuy-
Synthes and Acumed) versus a straight plate. To the
best of our knowledge this is the rst study that
evaluates the prominence of several different types
of clavicular plates using a 3D-CT-scan reconstruc-
tion analysis and an automated plate-to-bone align-
ment. Superior plate xation of the clavicle presents
several unique demands, due to the complex, highly
variable, bony architecture of the clavicle and its
immediate subcutaneous location (16). Several types
of plates have been used to x the broken clavi-
cle (21,25,27). The use of pelvic reconstruction plates
RESULTS
In 65 out of 728 cases a bad t was observed
(Table I). There are signicant more bad ts with
the straight plates compared to the precontoured
plates in group A (p < 0.001), B (p = 0.004) and C
(p < 0.001). There is no statistical difference be-
tween the number of bad ts between precontoured
plates in group A (p = 1.000) and C (p = 0.695).
In 663 cases the bone-plate distance could be
measured. The mean bone-plate distance of the dif-
ferent plates can be seen in Table I. The mean bone-
plate distance is signicant higher with the straight
plates compared to the precontoured plates in group
A (P < 0.001), B (p < 0.001) and C (p < 0.001). The
mean bone-plate distance is signicant higher with
the precontoured DePuy-Synthes plates compared
to the precontoured Acumed plates in group A
(p < 0.001) and in group C (p < 0.001).
Fig. 4. — Measurement of plate-bone distance and prominence
Fig. 3. — Optimalization of the plate-bone t
van tongel-.indd 304 29/09/14 10:43

Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
eValuaTion of prominence of sTraigHT plaTes anD precounTereD claVicle plaTes 305
To our knowlegde only two studies evaluated
the feasibility of clavicular osteosynthesis (7,8).
Grechting et al xed manually 4 different AO lock-
ing compression plates on 49 different clavicles.
They positioned manually the plate on cadavers in
an optimal surgical way on the superior surface of
the clavicle (7). They dened a good t of the plate
has been proposed because they are easier to con-
tour and can provide a better plate-bone contact
then non-contoured locking plates. But contouring
is time consuming and reconstruction plates are me-
chanically weaker then angularly stable im-
plants (4,7,19,21,26). This is the reason why we opted
to study only angular stable implants.
Fig. 5. — Box-plot of the prominence of the different plates
Table I. — Number of bad fits and measurement of plate-bone distance and prominence
plate bad fit mean mean plate-bone distance
(+/- SD)
mean maximum plate bone
distance (+/- SD)
mean prominence (+/-SD)
S1 13 1,26 (+/- 0,29) 4,22 (+/- 1) 6,22 (+/- 0,91)
S2 11 1,42 (+/- 0,38) 4,98 (+/- 1,36) 7,28 (+/- 1,4)
S3 15 1,63 (+/- 0,42) 5,51 (+/- 1,39) 7,66 (+/- 1,43)
DS1 1 1,12 (+/- 0,36) 3,48 (+/- 1,15) 5,11 (+/- 1,08)
DS2 1 1,13 (+/- 0,36) 3,51 (+/- 0,95) 5,20 (+/- 0,84)
DS3 4 1,24 (+/- 0,35) 3,93 (+/- 1,23) 5,49 (+/- 0,97)
A1 1 0,98 (+/- 0,32) 3,40 (+/- 1,1) 5,21 (+/- 0,82)
A2 2 1,02 (+/- 0,55) 3,20 (+/- 1,42) 5,25 (+/- 1,19)
A3 3 0,93 (+/- 0,29) 3,19 (+/- 0,96) 5,13 (+/-0,84)
A4 5 1,08 (+/- 0,24) 3,79 (+/- 0,74) 5,70 (+/-0,62)
A6 0 0,75 (+/-0,27) 2,41 (+/- 0,93) 4,58 (+/- 0,62)
A7 2 0,93 (+/- 0,29) 3,17 (+/- 1,07) 5,12 (+/- 0,82)
A8 2 0,94 (+/- 0,38) 3,01 (+/- 1,22) 4,98 (+/- 0,93)
van tongel-.indd 305 29/09/14 10:43

306 a. Van Tongel, T. Huysmans, b. amiT, j. sijbers, f. Vanglabbeek, l. De WilDe
Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
in the fact that the length of the Acumed plates is
shorter than de DePuy-Synthes plate. As stated in
the methods, we compared plates with the same
numbers of holes and not the length because we
think that during surgery the surgeon will be more
guided by the number of holes then by the length of
the plate.
Prominence of the plate can be a concern because
this may lead to irritation of the soft tissues around
the plate and as it is the most common reason for
reintervention after clavicular plate osteosynthesis.
This is the reason why we also studied the largest
minimum distance between the plate and the bone
(also taking the beveled edges into account). There
was a statitiscal difference between the straight
plates and the precontoured plates. The mean
hardware prominence in straight plates is 7 mm
and in pre contoured plates 5.2 mm. From clincical
point of view this 1.8 mm difference is probably
relevant, knowing that a thin periost, platsyma
and the skin only cover the superior part of the
clavicle. It has been described that the normal
thickness of myocutaneous platsyma ap is
2.2 mm (1). When evaluating the difference between
Acumed plates and the DePuy-Synthes plates in
the group with 8 holes, there is no signicant
difference which can be explained because the
thickness of the DePuy-Synthes plate is less. The
mean difference is also only 0,3 mm and probably
clinical not relevant.
There are some weaknesses in this study. First we
stimulated a transverse fracture in the middle of the
clavicle. We did not take any comminution or
different location of the fracture in the shaft into
account and a perfect anatomical reduction was
always considered as the ultimate surgical goal. A
non perfect anatomical reduction, thus a reconstruc-
tion of the bone to the plane rather than vice versa,
can probably inuence both the prominence and the
location. Second, we did not take the soft-tissue
envelop around the plate into account, which can
signicantly inuence the likelihood to provoke
irritation.
Third, hardware irritation is still a subjective
feeling and in this study is not possible to analyse
the correlation between hardware prominence and
the patients complaint.
if three screws could be safely applied through
either side of a mid-shaft fracture. In case of com-
minution, or a buttery fragment, two screws were
also accepted. Huang et al used axial radiographs
of 200 clavicles. Digitized representations of the
3 precontoured Acumed plates were freely trans-
lated and rotated along each clavicle to determine
the quality of t and the location of “best t.” (8).
“Best t” was dened as placement of the plate in a
location that “best” matched the S-shaped curvature
of the clavicle with minimum anterior or posterior
plate overhang. Both methods, the clinical or radio-
logical evaluation are prone to visual bias which is
overcome using the fully automated technique of
‘the best t’ to bone-plate alignment. To obtain an
optimal reproduction of a real life situation, we en-
sured at least three bicortical screws on each side of
the fracture with a minimum distance of 4 mm from
the fracture line, and ensure that each screw is
surrounded by enough bone (at least 3 mm). We
dened this seen as the worst-case scenario. As
described there are statistical signicant (p < 0.001)
less bad ts with the precontoured plates compared
to the straight plates and this as well for 6,7 and
8 holes and not for both groups of precontoured
plates. Nevertheless, in a clinical setting these
straight plates can be useful for the surgeon because
screws can be inserted in an oblique way and not in
the pre-determined locking direction.
We also measured, without any human bias, the
distance between the bone and the plate in a three-
dimensional way. The mean maximum plate-bone
distance with straight plates but also with precon-
toured plates is larger than compared to the results
of Grechting et al. In our opinion this is can be ex-
plained by the different measurement techniques
that are used (two-dimensional versus three-dimen-
sional technique). This means that the study of
Grechting measured a projection of the real length.
Also we do not fully understand how an accuracy of
0.1 mm can be obtained clinically.
In our study, the mean maximum distance with
the precontoured plates is signicant lower than the
straight plates. Next, the mean maximum distance
of the Acumed plates is signicant lower for to two
subgroups (6 holes, 8 holes) compared to the DePuy-
Synthes plate. A possible explanation can be found
van tongel-.indd 306 29/09/14 10:43

Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
eValuaTion of prominence of sTraigHT plaTes anD precounTereD claVicle plaTes 307
11. Kulshrestha V, Roy T, Audige L. Operative versus non-
operative management of displaced midshaft clavicle frac-
tures : a prospective cohort study. J Orthop Trauma 2011 ;
25 : 31-38.
12. Liu HH, Chang CH, Chia WT, Chen CH, Tarng YW,
Wong CY. Comparison of plates versus intramedullary
nails for xation of displaced midshaft clavicular fractures.
J Trauma 2010 ; 69 : E82-87.
13. Lorensen WE, Cline HE. Marching cubes : A high resolu-
tion 3D surface construction algorithm. SIGGRAPH Com-
put Graph 1987 ; 21 : 163-169.
14. McKee RC, Whelan DB, Schemitsch EH, McKee MD.
Operative versus nonoperative care of displaced midshaft
clavicular fractures : a meta-analysis of randomized clinical
trials. J Bone Joint Surg Am 2012 ; 94 : 675-684.
15. Millett PJ, Hurst JM, Horan MP, Hawkins RJ.
Complications of clavicle fractures treated with intra-
medullary xation. J Shoulder Elbow Surg 2011 ; 20 : 86-
91.
16. Mullaji AB, Jupiter JB. Low-contact dynamic compres-
sion plating of the clavicle. Injury 1994 ; 25 : 41-45.
17. Nordqvist A, Petersson C. The incidence of fractures of
the clavicle. Clin Orthop Relat Res 1994 ; 127-132.
18. Postacchini F, Gumina S, De Santis P, Albo F. Epidemi-
ology of clavicle fractures. J Shoulder Elbow Surg 2002 ;
11 : 452-456.
19. Robertson C, Celestre P, Mahar A, Schwartz A. Recon-
struction plates for stabilization of mid-shaft clavicle frac-
tures : differences between nonlocked and locked plates in
two different positions. J Shoulder Elbow Surg 2009 ; 18 :
204-209.
20. Russo R, Visconti V, Lorini S, Lombardi LV. Displaced
comminuted midshaft clavicle fractures : use of Mennen
plate xation system. J Trauma 2007 ; 63 : 951-954.
21. Shen JW, Tong PJ, Qu HB. A three-dimensional recon-
struction plate for displaced midshaft fractures of the clavi-
cle. J Bone Joint Surg Br 2008 ; 90 : 1495-1498.
22. Smekal V, Irenberger A, Struve P, Wambacher M,
Krappinger D, Kralinger FS. Elastic stable intramedul-
lary nailing versus nonoperative treatment of displaced
midshaft clavicular fractures-a randomized, controlled,
clinical trial. J Orthop Trauma 2009 ; 23 : 106-112.
23. Thyagarajan D, Day M, Dent C, Williams R, Evans R.
Treatment of mid-shaft clavicle fractures : A comparative
study. Int J Shoulder Surg 2009 ; 3 : 23-27.
24. van der Meijden OA, Gaskill TR, Millett PJ. Treatment
of clavicle fractures : current concepts review. J Shoulder
Elbow Surg 2012 ; 21 : 423-429.
25. VanBeek C, Boselli KJ, Cadet ER, Ahmad CS,
Levine WN. Precontoured plating of clavicle fractures :
decreased hardware-related complications ? Clin Orthop
Relat Res 2011 ; 469 : 3337-3343.
26. Wagner M. General principles for the clinical use of the
LCP. Injury 2003 ; 34 Suppl 2 : B31-42.
27. Werner SD, Reed J, Hanson T, Jaeblon T. Anatomic
relationships after instrumentation of the midshaft clavicle
CONCLUSIONS
To conclude precontoured plates of the clavicula
diminish signicantly the hardware prominence.
There exists a difference in hardware prominence
between different brands of precontoured plates but
this difference is limited and in most cases not sig-
nicant. The studied precontoured plates are suf-
ciently anatomically curved and can cover the big
variety of curves of the clavicle.
REFERENCES
1. Bauer T, Schoeller T, Rhomberg M, Piza-Katzer H,
Wechselberger G. Myocutaneous Platysma Flap for Full-
Thickness Reconstruction of the Upper and Lower Lip and
Commissura. Plastic and Reconstructive Surgery 2001 ;
108 : 1700-1703.
2. Bostman O, Manninen M, Pihlajamaki H. Complica-
tions of plate xation in fresh displaced midclavicular frac-
tures. J Trauma 1997 ; 43 : 778-783.
3. Chen QY, Kou DQ, Cheng XJ, Zhang W, Wang W,
Lin ZQ, Cheng SW, Shen Y, Ying XZ, Peng L, Lv CZ.
Intramedullary nailing of clavicular midshaft fractures in
adults using titanium elastic nail. Chin J Traumatol 2011 ;
14 : 269-276.
4. Demirhan M, Bilsel K, Atalar AC, Bozdag E,
Sunbuloglu E, Kale A. Biomechanical comparison of xa-
tion techniques in midshaft clavicular fractures. J Orthop
Trauma 2011 ; 25 : 272-278.
5. Ferran NA, Hodgson P, Vannet N, Williams R,
Evans RO. Locked intramedullary xation vs plating for
displaced and shortened mid-shaft clavicle fractures : a ran-
domized clinical trial. J Shoulder Elbow Surg 2010 ; 19 :
783-789.
6. Gerber C, Pennington SD, Lingenfelter EJ,
Sukthankar A. Reverse Delta-III total shoulder replace-
ment combined with latissimus dorsi transfer. A prelimi-
nary report. J Bone Joint Surg Am 2007 ; 89 : 940-947.
7. Grechenig W, Heidari N, Leitgoeb O, Prager W,
Pichler W, Weinberg AM. Is plating of mid-shaft clavicu-
lar fractures possible with a conventional straight 3.5 mil-
limeter locking compression plate ? Acta Orthop Trauma-
tol Turc 2011 ; 45 : 115-119.
8. Huang JI, Toogood P, Chen MR, Wilber JH,
Cooperman DR. Clavicular anatomy and the applicability
of precontoured plates. J Bone Joint Surg Am 2007 ; 89 :
2260-2265.
9. Huysmans T, Sijbers J, Verdonk B. Automatic construc-
tion of correspondences for tubular surfaces. IEEE Trans
Pattern Anal Mach Intell 2010 ; 32 : 636-651.
10. Huysmans T, Sijbers J, Verdonk B. Parameterization of
tubular surfaces on the cylinder. Journal of the Winter
School of Computer Graphics 2005 ; 13 : 97-104.
van tongel-.indd 307 29/09/14 10:43

308 a. Van Tongel, T. Huysmans, b. amiT, j. sijbers, f. Vanglabbeek, l. De WilDe
Acta Orthopædica Belgica, Vol. 80 - 3 - 2014
complications of plate xation of clavicle fractures. Arch
Orthop Trauma Surg 2012 ; 132 : 617-625.
30. Zlowodzki M, Zelle BA, Cole PA, Jeray K, McKee MD.
Treatment of acute midshaft clavicle fractures : systematic
review of 2144 fractures : on behalf of the Evidence-Based
Orthopaedic Trauma Working Group. J Orthop Trauma
2005 ; 19 : 504-507.
with 3.5-mm reconstruction plating : an anatomic study.
J Orthop Trauma 2011 ; 25 : 657-660.
28. Wijdicks FJ, Houwert RM, Millett PJ, Verleisdonk EJ,
Van der Meijden OA. Systematic review of complications
after intramedullary xation for displaced midshaft clavicle
fractures. Can J Surg 2013 ; 56 : 58-64.
29. Wijdicks FJ, Van der Meijden OA, Millett PJ,
Verleisdonk EJ, Houwert RM. Systematic review of the
van tongel-.indd 308 29/09/14 10:43