I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
155–163
USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC
INSTRUMENTS IN CLINICAL PRAXIS
UPORABA PACIENTU PRILAGOJENIH INSTRUMENTOV,
IZDELANIH Z DODAJALNO IZDELAVO, V KLINI^NI PRAKSI
Igor Drstven{ek
1
, Ur{ka Kostev{ek
1
, Toma` Brajlih
1
, Nata{a Ihan Hren
2
, Matja` Merc
3
,
Toma` Toma`i~
4
, Sne`ana Stevi~
5
, Matja` Vogrin
3
, Andrej Moli~nik
3
1
University of Maribor, Faculty of Mechanical Engineering, Smetanova ulica 17, 2000 Maribor, Slovenia
2
University of Ljubljana, Faculty of Medicine, Vrazov trg 2, 1000 Ljubljana, Slovenia
3
University Medical Centre Maribor, Ljubljanska ulica 5, 2000 Maribor, Slovenia
4
General Hospital Murska Sobota, Ulica dr. Vrbnjaka 6, 9000 Murska Sobota, Slovenia
5
University of Pri{tina-Kosovska Mitrovica, Medical Faculty, Filipa Vi{nji~a bb, 38220 Kosovska Mitrovica, Serbia
Prejem rokopisa – received: 2018-07-15; sprejem za objavo – accepted for publication: 2018-10-18
doi:10.17222/mit.2018.152
Using a combination of Computer-Assisted Design (CAD), expert surgical knowledge and Additive-Manufacturing (AM)
technologies, it is nowadays possible to offer patients an individualized treatment that is better planned, more predictable and
more reliable than traditional surgical procedures. Regardless of the specifics of different medical fields, planning is common to
all of them when it comes to an invasive intervention into the human body. Surgical planning is always based on diagnostic data
that need to be gathered, presented and evaluated in a way that best suits the final purpose. To understand the general needs of
the surgical planning process, several surgical cases have been explored and analyzed in the Additive Manufacturing Laboratory
(AML) at the Faculty of Mechanical Engineering, University of Maribor over the past 10 years. The common denominator in all
cases is the reconstruction of diagnostic data into a 3D model that is later used for planning and is, after the confirmation of the
planned therapeutic parameters, transformed into tangible surgical equipment by means of AM. The 3D planning process has a
lot of benefits but requires some skills uncommon among the medical experts. The results of the presented research summarize
the particularities of medical 3D planning and provide guidelines for the wider adoption of medical 3D planning and the use of
additively manufactured patient-specific instruments.
Keywords: additive manufacturing in medicine, 3D medical planning, surgical guides, patient-specific instruments and implants
Uporaba ra~unalni{ko podprtega konstruiranja, specialisti~nih kirur{kih znanj in tehnologij dodajalne izdelave nudi pacientom
individualno, bolje na~rtovano, predvidljivej{o in zanesljivej{o obravnavo kot tradicionalna kirurgija. Natan~no na~rtovanje
operacij je skupno vsem medicinskim podro~jem, kadar gre za invaziven poseg v ~lovekovo telo, ne glede na medicinsko
specializacijo. Na~rtovanje kirur{kih posegov vedno temelji na diagnosti~nih podatkih, ki jih morajo zajeti, predstaviti in
ovrednotiti na na~in, ki ustreza kon~nemu cilju. Da bi razumeli splo{ne potrebe na~rtovanja kirur{kih posegov, so v zadnjih 10
letih v Laboratoriju za dodajalno izdelavo Fakultete za strojni{tvo Univerze v Mariboru raziskali in analizirali precej kirur{kih
primerov. Vsem primerom je skupna rekonstrukcija diagnosti~nih podatkov v trirazse`ne modele, ki jih kasneje uporabijo za
na~rtovanje, po potrditvi na~rtovanih terapevtskih parametrov pa preoblikujejo v otipljiv kirur{ki pripomo~ek s pomo~jo
dodajalne izdelave. Proces trirazse`nega na~rtovanja ima mnogo prednosti, zahteva pa nekaj znanj in spretnosti, ki med
zdravniki niso obi~ajne. Rezultati prikazanih raziskav povzemajo posebnosti trirazse`nega na~rtovanja in podajajo smernice za
bolj raz{irjeno uporabo pacientom prilagojenih medicinskih pripomo~kov, narejenih s tehnologijami dodajalne izdelave.
Klju~ne besede: dodajalna izdelava v medicini, trirazse`no medicinsko na~rtovanje, kirur{ka vodila, pacientu prilagojeni
instrumenti in vsadki
1 INTRODUCTION
In recent decades, additive manufacturing (AM) has
offered help in medical reconstruction and planning
procedures to assist surgeons in re-establishing the
functionality of injured or otherwise affected body parts
in their patients. The first cases of AM-supported surgery
in the Additive Manufacturing Laboratory (AML) at the
Faculty of Mechanical Engineering, University of
Maribor, covered the area of cranial and maxillofacial
surgeries for which implants were custom-made for each
patient. The first orthopaedic cases soon followed with
bespoke resection guides that enabled the precise and
faster replacement of shoulder, hip and knee joints. A
similar technique has been used to manufacture patient-
specific drill-guide templates that enabled optimal pe-
dicle screw placement. The method has been evaluated
by performing a clinical study involving the manufacture
of templates for the lumbar and sacral regions that
enabled simultaneous multiple-level screw implanting.
All the mentioned surgical cases come from different
areas and have different motivations for interdisciplinary
cooperation during their performance. Yet, they share a
common source of planning data and an engineering
approach independent of the particularities of the me-
dical field. All 3D planning data used in such a process
are gathered through a series of 3D imagings performed
using one of the applicable techniques (MRI, X-ray CT,
Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163 155
UDK 617:004.9-045.43:616-76 ISSN 1580-2949
Original scientific article/Izvirni znanstveni ~lanek MTAEC9, 53(2)155(2019)
*Corresponding author e-mail:
igor.drstvensek@um.si

CBCT, USI). These data usually come in the form of
DICOM that needs to be converted into 3D, CAD data,
suitable for further manipulation. The main problem of
reliability is the accuracy of the gathered data, which is
affected by the technique itself, the diagnosed body part
and the operator or human factors at the time of imaging.
Defects in the craniofacial skeleton are of either con-
genital (birth defects), developmental orthognathic
deformities) or accidental (resulting from trauma, infec-
tion, tumour, etc.) cause. The purpose of reconstructing
abnormalities is primary functional. In cranial-maxillo-
facial treatments, implants need to fulfil an aesthetic
function too. The possibilities for their prefabrication by
means of serial production are very limited. Therefore,
these implants and instruments are individually made
using either subtractive (CNC milling) or additive (Laser
sintering or melting) technologies. The production of
such implants starts by reconstructing a three-dimen-
sional data set of the problematic area (skull, mandibular
area, pelvis, shoulder rim, etc.) from CT or MRI
two-dimensional pictures. This reconstructed model is
then used for modelling the missing bones. The digital
model is then manufactured using either CNC milling
(PEEK and ceramic – ZrO
2
,A L
2
O
3
– implants) or
Additive Manufacturing (metal implants and surgical
instruments).
Besides producing a patient-specific implant, the 3D
model reconstructed from diagnosing imagery can be
effectively used for the preparation of the surgical
treatment. This is especially important in cases where the
functionality of the body part must be re-established.
Such examples are orthopaedic surgeries where the
functionality of the limbs after the surgical procedure
depends on the positioning of the implant. Virtual
models can be used to study the surgical procedures, like
the directions of implantation, the required pre-oper-
ational treatments and the preparations, etc. Very
interesting in this sense are total hip replacement (THR)
operations that are, due to an ageing society and
unhealthy lifestyles, constantly on the increase. Since the
primary THRs are performed on ever-younger patients
the number of revisions is constantly increasing too.
1
Two-dimensional pre-operative planning,
2,3
which is due
to a simple process still in general use, has several
disadvantages that often lead to longer surgical pro-
cedures,
4
inaccurate positioning of the endoprosthesis,
5
additional bone loss,
6
limb length discrepancy,
7
longer
rehabilitation and increased overall costs.
8,9
This observation also proved itself in a case of the
use of pedicle screw systems for vertebral fusion. It is
routinely used for stabilization between two or more
spinal levels in degenerative, traumatic, oncogenic or
pre-operatively deformed vertebrae.
10
During pedicle
screw insertion, it is important both to select the correct
size of the screw and to place it properly within the
pedicle to ensure good anchoring. The free-hand tech-
nique has a high associated rate of unplanned per-
foration, which is the major specific complication of
pedicle screw placement and causes a high risk of bone
weakening or lesions of the spinal cord, nerve roots or
blood vessels.
11
Individualized drill-guide templates for
the lumbar spine were first introduced in 1998. The
recent study of the AML and University Clinical Centre
in Maribor established that the use of a multi-level drill
guide improves the accuracy of pedicle screw positioning
in the lumbar and sacral spine, and that the application of
multi-level drill guides is reasonable in screw placement
on the first sacral level.
12
2 CLINICAL EXPERIENCES
2.1 Custom-made cranio-maxillofacial implants
In the past 12 years, 31 cases of cranioplasty have
been performed in a cooperation between the AML and
three clinical institutions in Slovenia. One of the patients
was treated with a titanium implant produced by SLM
technology and all the others by the indirect approach in
which the template of the cranial implant was additively
manufactured and later used to produce a silicone rubber
mould (SRM). A so-called bone cement, a mixture of
PMMA and ceramic powder, was moulded in the SRMs
to form a biocompatible implant.
The easiest way to reconstruct the structure of a
patient’s bones is to use those CT images that already
exist from previous treatments of the patient. A set of CT
images can be converted into a three-dimensional, digital
model using one of the available software packages, such
as: EBS (Ekliptik), Mimics (Materialise), 3D doctor
(Able Software), Amira (Thermo Fisher Scientific),
Dolphin 3D Surgery (Dolphin/Patterson Dental), 3D
Slicer (Open Source) or others.
13
The three-dimensional model can be further
manipulated with several CAD software packages. The
usual 3D modellers based on parametric, volume-
modelling techniques are not very well suited to the task.
Newer versions of these software packages (SolidWorks,
Delcam, etc.) enable the manipulation of triangulated
surface files, but using dedicated software, known from
Reverse Engineering fields, such as Magics (Mater-
ialise), PolyWorks (InnovMetric), MeshLab (open
source), MeshMixer (Autodesk) or others is much more
effective in terms of time and effort. Using these tools
the missing tissue can be modelled and saved as new
files. These can be further processed or used to produce
real implant models using one of the AM technologies.
Reconstructed models of the skull and the implant have
been manufactured using several different AM tech-
nologies.
In the case of skull implants Selective Laser Sintering
(SLS), PolyJet™ and stereolithography (SLA) were used
for the production of communication models and
patterns for implant production. SLS technology builds
parts from powder that is solidified in slices by a laser
beam. The powder is one of the well-known plastic
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
156 Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163

compounds (usually polyamide) stored in special con-
tainers beside the machine’s working surface. From here
it is applied by a "recoater" onto the working surface for
each layer, separately. PolyJet technology builds models
from photo polymeric resins. Each layer is jetted on the
work tray by a printing head and then cured by
ultraviolet light. The support material is later removed by
a water jet. Stereolithography produces parts in a vat
filled with a photopolymer resin. The resin is solidified
by means of a UV laser that selectively scans the surface
of the resin until the cross-section of a layer is solidified.
The support structure is made of the same material as the
part and needs to be manually removed after the process
is finished.
SLS produces rigid and resistant polyamide parts
with a relatively high surface roughness. The roughness
makes the SLS parts unsuitable for silicone rubber
moulding unless the parts undergo a surface treatment,
usually vibration polishing. The PolyJet process pro-
duces smooth parts with lower rigidity at a higher price.
On the other hand, the price difference in the case of
smaller parts such as the implant for cranioplasty is not a
substitute for PolyJet’s better performance in terms of
surface and dimensional quality. The parts produced with
stereolithography have comparable mechanical proper-
ties to the PolyJet parts, a slightly smoother surface, but
require a more demanding process for supporting
structure removal.
For the direct production of implants either Selective
Laser Melting (SLM) or Electron Beam Melting tech-
nology is used. Both can produce fully dense metal parts
and both provide the environment for producing a
biocompatible implant out of reactive metal alloys, such
as titanium grade 23. Both technologies employ a
powder-bed system with a recoater to deliver the
material onto the building tray. The SLM technology
uses a Yb-glass fibre laser and operates within a chamber
filled with argon with less than 0.8 % of remaining
oxygen. The EBM technology operates in vacuum with a
high-power electron beam (up to 4kW) that selectively
melts the powdered material and binds it into the final
product.
For the indirect production of biocompatible
implants, a modified SRM procedure has been used. A
SRM was made using a normal frame to hold the
silicone and the pattern.
14
The pattern holders were
purposely made from 5-mm steel wires to make some
room for the excess PMMA compound (Figure 1).
Since the material cannot be poured into the mould
through a normal gating the material needs to be mixed
in one half of the mould and then pressed together with
the other half to achieve the final form. Therefore, the
mould was equipped with guiding rods to prevent side
movements and with some extra openings to release the
excess amount of bone cement out of the mould during
the pressing process. (Figure 1). The finished implants
need to be sterilized by low-temperature sterilization to
avoid changes in the polymeric structure of the PMMA.
The cranioplasty operations have traditionally been
performed with no alterations to the standard procedure.
The implant was inserted into the skull of the patient and
fixed using titanium plates and screws. A CT inspection
showed good position of the bone cover. The whole
duration of the operation has been shortened by approxi-
mately 50 %, due to the preparation work (planning, fit
and function testing, production of custom-made im-
plant) carried out before the operation.
The 3D modelling of skull implants is a relatively
simple task because of the typical shape that can be
regenerated from the tissues adjacent to the defective
area. Other areas of the head are more demanding
because of the more complicated bone morphology. Such
is the case of a 24-year-old mentally healthy man, born
with hemifacial microsomia. This is a severe asymmetry
of the facial bone and soft tissues in the vertical, sagittal
and transverse plane combined with hearing impairment
on the affected side. He was treated using classic
orthognathic surgical procedures and by distraction
osteogenesis of the mandible. After these surgical
procedures the remaining defect of the bone and soft
tissues was partially compensated by a custom-made
titanium angular implant.
To find the optimal shape of the angular implant the
skull’s virtual model was split into two parts along the
nose plane. The healthy side was mirrored across the
nose plane and super-positioned over the left part of the
face.
The Boolean subtraction between the mirrored and
original skull part produced a reference model (Figure
2). The implant was modelled between the reference and
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163 157
Figure 1: a) Manufacturing of SRM mould, b) moulding of implant, c) molded PMMA implant with excess material at the parting plane
15

the original shape of the mandible. It had to be
lightweight and easy to handle during the operation.
To evaluate the implant’s shape models of the skull
and the implant were created using the SLS technology.
They enabled an effective communication between the
medical doctor and the engineer (Figure 3). After the
surgeon’s evaluation of the models the implant’s shape
was slightly changed to accommodate the mandibular
muscles. After confirming the final design, the implant
was produced out of Ti-grade 23 alloy, using SLM
technology.
2.2 Orthopedic surgical planning
In cases of orthopaedic surgery, the 3D planning and
surgical preparation play a very important role. The
implants in these cases come from serial production, but
their implantation depends on the skills of the surgeon
and on his/her planning possibilities. Conventional 2D
pre-operative planning is performed on the antero-
posterior (AP) pelvic radiography, which directly
provides the data about the inclination angle and the
centre of the hip’s rotation. The version angle can only
be calculated indirectly given that the lesser trochanter is
visible in the radiograph and that its size can be reliably
defined. This fact and the lack of reliable anatomical
landmarks accessible during the surgery make 2D
planning outdated and inappropriate for modern THA
operations.
16
Anatomical landmarks are intra-operational, easily
accessible bone structures that are clearly visible on the
radiographic images. These usually include the
medullary canal (shaft), the greater and lesser trochanter,
the acetabular roof, saddle and the teardrop (Figure 4a).
The problem of anatomical landmarks is that most of
them are only visible on the radiographic images (some
even because of overlapping bone structures), but most
of them cannot be seen “in vivo” during the surgery.
Mechanical references are the distances between the
different landmarks that define the functionality of the
hip joint. After the THA operation, the hip movement
must be restored into its anatomical state. This can only
be ensured by a proper definition of the mechanical
references, e.g., the original acetabular and femoral
rotation centre, the femoral and acetabular offset and the
leg length.
The measurements shown in Figure 4b are unreliable
because they depend on the circumstances of the 2D
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
158 Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163
Figure 2: Reference shape for implant design and the final implant’s shape
15
Figure 3: Angular implant: a) The model of the custom-made angular implant on the printed model of the patient’s head skeleton, b) the Ti64
implant after SLM, c) the Ti64 implant after polishing
15

radiograph’s acquisition, its magnification factor and the
influence of the version angles unidentifiable in the 2D
plane of the radiograph.
16
In a 3D space obtained from
the CT, datasets all the references can be measured
directly and their precision only depends on the accuracy
of the CT scanner.
To overcome the obstacles of 2D surgical planning its
benefits were combined with the possibilities of the 3D
space. This is obtained by reconstructing the bone
structures from the CT scans into 3D data. Prior to the
3D surgical planning process a coordinate system has to
be defined, which corresponds to the established surgical
practice. According to the proposition of Baauw M. et
al.
17
the Cartesian coordinate system of three planes
(sagittal, coronal and transversal) was defined, using 3D
landmarks found in the 3D model. The 3D planning
consists of two phases. In the first phase the anatomical
parameters of the hip joint are defined and in the second
phase the devices to transfer the defined parameters into
the patient’s body are modelled.
The anatomical parameters are the version and
inclination angles and the centre of rotation. In the 3D
space they can be precisely defined using geometrical
primitives and Boolean operators provided by the CAD
software package. By placing the primitives into the
virtual bone structure, the anatomical parameters of the
hip can be measured with an accuracy and reliability not
possible in the 2D planning process. Using Boolean
operators the bone structure was subtracted from the
pre-modelled, standard templates to model the custom-
made resection guides that can be produced by AM
technologies. These resection guides were used during
the surgery to help the surgeon maintain the calculated
anatomic parameters in the operational space (Figure 6).
A resection guide is an individual medical device made
of polyamide (PA) and intended solely for inter-
operational use to transfer the references of the virtual
coordinate system into the patient’s body, thus
facilitating the achievement of the planned parameters
(CR, INC, AV).
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163 159
Figure 4: a) Anatomical landmarks: 1. Femoral shaft; 2. Greater trochanter; 3. "Saddle"; 4. Lesser trochanter; 5. Acetabular roof; and 6. Tear-
drop, b) Mechanical references: 1. Hip rotation centre; 2. Longitudinal axis of the proximal femur; 3. Femoral offset; 4. Acetabular offset; 5. Hip
length ("Leg length discrepancy" – distances between the 6L and 6R)
16
Figure 5: Individual medical devices for the femoral resection
15

The described method has been clinically tested on
five patients. Informed consent was obtained. All the
patients suffered from primary osteoarthritis type IIA
according to the Paprosky classification system. The
post-operative CT analysis was performed to establish
the quality of planned parameters transfer with the used
resection guides.
2.3 Drill-guide templates for positioning of pedicle
screws
Using a multi-level template for positioning the
pedicular screws into multiple vertebrae levels strongly
depends on the ability to put the spine into the same
intravertebral positions as during the diagnostic imaging.
To achieve this condition, a CT scan of the lumbar and
sacral spine was performed on patients lying in the prone
position to simulate similar facet joint relations as during
the operating procedure. The images were transferred to
a workstation running EBS ver. 2.2.1 (Ekliptik, Slovenia)
software to generate a 3D reconstruction model for the
targeted lumbar or sacral vertebra. The 3D spine model
was used to determine the optimal screw size and
orientation. The pedicle circumference and direction
were investigated by pointing the virtual screw towards
the centre of each half of the corpus vertebrae. Then, a
3D vertebral model was reconstructed with virtual
screws placed on both sides and on different levels. The
optimal screw length was also determined so that the end
of the screw reached 50–80 % of the vertebral diameter
(Figure 6).
Using the achieved virtual screws, a drill-guide
template was constructed with a surface designed to be
the inverse of the dorsal part of the facet joint. That was
meant to enable a lock-and-key mechanism fitting the
dorsal part of the facet to achieve minimal overlap. The
parts of the template for each pedicle screw were
connected to each other in the sagittal and transversal
plane to achieve the maximum stability of the template.
(Figure 8). Additionally, cylinders fitting a trajectory
hole were manufactured, allowing a temporary fixation
of the drill guide with K-wires. The drill guide was
produced with selective laser sintering technology from
Polyamide 12.
The drill-guide templates were clinically tested in an
open-label clinical trial on 20 subjects, featuring the
implantation of 54 pedicle screws using drill-guide
templates and 54 screws using the free-hand technique.
Fluoroscopy surveillance was used in both groups.
All the patients underwent a postoperative CT scan
that was used for the post-operative analysis of the screw
placements.
3 RESULTS
The cranial and angular implants were the first trials
that proved the usability of the CT scans for the accurate
design and production of the medical implants. The main
question in all cases was the reliability of the 3D data
reconstructed from the CT scans. The trials showed that
the achieved deviations fall within the dimensional field
of 1 mm and the main contribution to the final deviations
was in all cases added by the production process.
18
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
160 Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163
Figure 7: Design of a multi-level drill-guide template. A drill guide fits onto the dorsal elements of the facet joint. Parts of the guide are
connected to each other to achieve better angular stability.
15
Figure 6: Determination of screws’ direction. Screws are directed
through the centre of the pedicle towards the midpoint of the ipsi-
lateral half of the vertebral corpus.
15

The post-operative analysis of the positioned hip
implants using the described method was performed by
comparing the planned positioning parameters (INT, AV,
COR) with the achieved position of the implant assessed
from the post-operative CT scan. The results in Table 1
show that the largest deviation between the planned and
actual position occurs in the coronal plane (1.25 mm).
The absolute spatial difference is 1.40 mm. The
difference between the planned and the post-operative
inclination of the femur is relatively large, as compared
to the other orientation angles (2.27°). This is due to an
unreliable definition of the femoral anteversion that can
only be defined according to the direction of the
condyles in the knee joint, which are not visible in the
CT of the pelvis area. Therefore, the femoral anteversion
was measured against the position of the pelvis.
Table 1: Data of planned and post-operative centres of rotation, angles
of inclination and anteversion
Transversal
plane
Planned Post-operative Difference
CR_FE 95.82 mm 95.77 mm 0.05 mm
CR_AC 95.82 mm 95.50 mm 0.32 mm
Sagittal plane Planned Post-operative Difference
CR_FE 46.87 mm 46.28 mm 0.59 mm
CR_AC 46.87 mm 46.32 mm 0.55 mm
Coronal plane Planned Post-operative Difference
CR_FE 1.91 mm 2.97 mm 1.06 mm
CR_AC 1.91 mm 3.16 mm 1.25 mm
Inclination Planned Post-operative Difference
INC_FE 135 ° 132.73 ° 2.27 °
INC_AC 40.75 ° 41.53 ° 0.78 °
Anteversion Planned Post-operative Difference
AV_FE 15.13 ° 13.98 ° 1.15 °
AV_AC 26.73 ° 24.77 ° 1.96 °
The usability and accuracy of the drilling templates
for the pedicle screw placement were established in two
ways. Firstly, the post-operative positions of the screws
were compared to the planned positions by reconstruct-
ing the 3D models of the positioned screws from the
post-operative CT scan and super-positioning them over
the planned 3D models, as shown in Figure 8.
The comparison in most cases showed negligible
deviations. A more insightful analysis was performed by
comparing the screw accuracy and the pedicle cortex
perforation (level of pedicle violation) between the
control and the template group. The accuracy was
measured using EBS ver. 2.2.1 (Ekliptik, Slovenia)
software and defined as the eccentric position of the
screw according to the smallest diameter of the ellipse of
the pedicle.
In both groups 54 pedicle screws were inserted. The
perforation of the pedicle cortex or the cortex of the
corpus vertebrae was significantly lower in the drill-
guide group. The screws in the control group were
displaced significantly more medially from the centre of
the pedicle and directed more laterally from the optimal
trajectory in the sagittal plane. There was no statistically
important difference between the groups in the
transversal plane, although the SD level in the control
group was higher. The screw length violation occurred
less frequently in the drill-guide group; but without the
difference being significant (Table 2).
Table 2: Pedicle screw position measurements for lumbar and 1
st
sac-
ral level. SD standard deviation,
a
Negative value means displacement/
deviation laterally,
b
Negative value means displacement/deviation
caudally,
c
The tip of the screw exceeds 50–80 % of the corpus
vertebrae diameter.
Drill guide Control p-value
No. of screws 54 54
Perforation of cortex 6 21 <0.001
Displacement sagittal
(mm), mean (SD)
a
0.3 (3.4) 1.5 (3.2) 0.05
Deviation sagittal (°),
mean (SD)
a
–1 (5) –6 (8) <0.001
Displacement transversal
(mm), mean (SD)
b
–0.7 (1.5) –0.2 (2.6) 0.28
Deviation transversal (°),
mean (SD)
b
–1 (5) 0 (11) 0.71
Screw length violation
c
14 20 0.21
4 DISCUSSION
The presented results clearly show the usability and
benefits of the 3D planning and the guided surgical
principles. Benefits and tremendous time savings were
observed when using custom-made cranial implants. In
the cases of maxillofacial operations, the use of custom-
made implants shows the same potential, but the time
savings are not so obvious, because of the more com-
plicated surgical process. But in any case, the surgeon
has a much better insight into the surgical procedure due
to the 3D planning process and the communication
models produced during the planning.
The benefits of 3D planning are even greater in
orthopaedic surgical planning, because of the better
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163 161
Figure 8: Comparison of the planned and the actual positions of the
pedicle screws
15

possibilities to assess all the anatomical/geometrical data
of the inspected body part. The results of the placement
of five primary endoprostheses show that the deviations
between the planned and achieved positions of the
implants are almost negligible. This means that the
planned, anatomical kinematics of the hip was
re-established and gives a great deal of confidence that
the endoprostheses will bear the planned stresses and
survive the planned life-cycle.
The results of the use of the drill templates showed
that a manufactured template significantly reduces the
perforation risk of pedicle or vertebral corpus cortex in
comparison with the free-hand technique under fluoro-
scopy supervision, which is more accurate than the
classic free-hand technique. Due to the lower perforation
risk, the probability of neural or vascular lesions was
also reduced.
11
When applying the template, the screw
was more precisely centred in the pedicle, especially in
the sagittal plane, resulting in a better stability of the
screw and a higher pullout force of the screw. During the
operating procedure, eliminating any tilting of the
template was the most challenging part, especially in the
transversal plane. This was minimized by the precise
striping of the soft tissue lying next to the facet joints
and applying moderate pressure to the template when
fixing it and introducing a screw.
The same problem can be observed in THR
operations, where the bones needed to be cleaned to
provide reliable fitting between the bone and the
resection guide. For this reason, the use of a resection
guide requires the precise preparation of soft tissues next
to the attaching areas and precise fitting and fixing of the
guide. When this is achieved, a guide can be easily used,
resulting in sufficient accuracy.
In cases of resection guides for orthopaedic surgery
and drill guides for the pedicle screw placement, the
accuracy of the overall process plays an important role.
Again, it was established that the overall process with
final plastic (either PA12 or photocured acrylic plastic)
parts ensures the required accuracy for the clinical use of
the guides.
5 CONCLUSIONS
The presented cases show that the use of modern
technologies combined with interdisciplinary coopera-
tion improves the overall effectiveness of surgeons,
shortens the procedures, and lowers the risks and costs of
the health-insurance system, providing more accurate
and reliable planning and surgical processes.
The 3D reconstruction of bone tissues from the CT
scanning data, and 3D pre-operational planning, make
the surgical process more accurate and more reliable, but
require better preparation of the referencing tissues. The
accuracy of positioning the instrument is ensured by
patient-specific templates designed to fit perfectly to the
morphology of the patient’s bones. In cases where only
the implant is custom made for the patient and the
implantation method remains unaltered, the surgical
process can be significantly shorter, compared to serially
made implants, mainly due to the pre-operational
planning and the perfect fit of the implants. The use of
patient-specific instruments such as drilling or resection
guides requires an alteration to the standard surgical
process. This involves the better preparation of hard
tissues, the introduction of new, patient-specific instru-
ments and the adaptation of the operating field to them.
The introduction of Additive Manufacturing and 3D
planning into the surgical process requires the intro-
duction of engineering personnel into the surgical team,
which needs to be foreseen in the organizational scheme
of the institution. All these represent obstacles to a wider
adoption of new technologies in medical/surgical fields
and keeps these kinds of approaches at the level of pilot
projects. Nevertheless, the presented results show that
AM has huge potential in medicine, mainly due to its
particularities that are in line with the requirements of
additive technologies, e.g., one-off production, compli-
cated shapes, rapid development, etc.
Acknowledgement
Parts of the present research were financially
supported by the Slovenian Research Agency (research
core funding no. P2-0157). The authors would like to
thank the Department of Radiology in UCC Maribor,
Slovenia, for assistance with the CT data scanning and
data preparation.
6 REFERENCES
1
V. Leva{i~, V. Pi{ot, I. Milo{ev, Arthroplasty Register of the
Valdoltra Orthopaedic Hospital and implant retrieval program, Zdrav.
Vestn., 78 (2009), 73–80, doi:10.6016/451
2
A. Huppertz, S. Radmer, M. Wagner, T. Roessler, B. Hamm, M.
Sparmann M, Computed tomography for preoperative planning in
total hip arthroplasty: what radiologists need to know, Skeletal
Radiol., 43 (2014) 1041–1051, doi:10.1007/s00256-014-1853-2
3
C. Zeng, W. Xing, Z. Wud, H. Huang, W. Huang W, A combination
of three-dimensional printing and computer-assisted virtual surgical
procedure for preoperative planning of acetabular fracture reduction,
Injury., 47 (2016), 2223–2227. doi:10.1016/j.injury.2016.03.015
4
H. R. Blackley, Planning and management of the difficult primary
hip replacement: preoperative planning and technical considerations,
Instr Course Lect, 49 (2000), 3–11
5
J. P. McAuley, Preoperative planning to prevent dislocation of the
hip, Orthop Clin North Am, 32 (2001), 579–586, doi:10.5435/
00124635-200511000-00005
6
F. S. Haddad, Femoral bone loss in total hip arthroplasty: classifi-
cation and preoperative planning, Instr Course Lect., 49 (2000),
83–96, doi:10.2106/00004623-199910000-00013
7
K. T. Suh, Comparison of preoperative templating with postoperative
assessment in cementless total hip arthroplasty, Acta Orthop
Scand.,75 (2004), 40–44, doi:10.1080/00016470410001708070
8
O. Cech, Preoperative planning and surgical technic in achieving
stability and leg length equality in total hip joint arthroplasty, Acta
Chir Orthop Traumatol Cech, 69 (2002) 362–8
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
162 Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163

9
WM. Goldstein, Leg length inequality in total hip arthroplasty,
Orthopedics, 28 (2005) 9, 336–341, doi:10.1007/s12178-013-9180-0
10
T. A. Mattei, M. S. Meneses, J. B. Milano, R. Ramina, "Freehand"
technique for thoracolumbar pedicle screw instrumentation: critical
appraisal of current “state-of-art”, Neurol India, 57 (2009)6 ,
715–721, doi:10.4172/2165-7939.S6-006
11
S. C. Zeiller, J. Lee, M. Lim, A. R. Vaccaro, Posterior thoracic
segmental pedicle screw instrumentation: evolving methods of safe
and effective placement, Neurol India, 53 (2005), 458–465
12
M. Merc, I. Drstven{ek, M. Vogrin, T. Brajlih, G. Re~nik, A
multi-level rapid prototyping drill guide template reduces the
perforation risk of pedicle screw placement in the lumbar and sacral
spine, Archives of orthopaedic and trauma surgery, 133 (2013)7 ,
893–899, doi:10.1007/s00402-013-1755-0
13
T. Wohlers, Wohlers Report, Fort Collins, 2018
14
H. Rotaru, M. Baciut, H. Stan, S. Bran, H. Chezan, A. Iosif, M.
Tomescu, SG. Kim, A. Rotaru, G. Baciut, Silicone rubber mould cast
poly-ethyl-methacrylate-hydroxyapatite plate used for repairing a
large skull defect, J Craniomaxillofac Surg, 34 (2006) 4, 242–246,
doi:10.1016/j.jcms.2006.01.005
15
Courtesy of Additive Manufacturing Laboratory, Faculty of
Mechanical Engineering, University of Maribor
16
T. Scheerlinck, Primary hip arthroplasty templating on standard
radiographs A stepwise approach, Acta Orthop. Belg., 76 (2010),
432–442
17
M. Baauw, G. G. van Hellemondt, L. M. van Hooff, M. Spruit, The
accuracy of positioning of a custommade implant within a large
acetabular defect at revision arthroplasty of the hip, Bone Joint J.,
97-B (2015) 6, 780–785, doi:10.1302/0301-620X.97B6.35129
18
I. Drstven{ek, N. Ihan Hren, T. Strojnik, T. Brajlih, B. Valentan, V.
Poga~ar, T. Zupan~i~ Hartner, Applications of rapid prototyping in
cranio-maxilofacial surgery procedures, International journal of
biology and biomedical engineering, 2 (2009) 1, 29–38
I. DRSTVEN[EK et al.: USE OF ADDITIVELY MANUFACTURED PATIENT-SPECIFIC INSTRUMENTS IN CLINICAL PRAXIS
Materiali in tehnologije / Materials and technology 53 (2019) 2, 155–163 163