European Journal of Dental Education ISSN 1396-5883

A novel anatomical ceramic root canal simulator for

endodontic training

cavin1, J.-C. Hornez5 and E. Deveaux1
L. Robberecht1,2,5, F. Chai2, M. Dehurtevent3, P. Marchandise4, T. Be
1
Department of Restorative Dentistry and Endodontics, Faculty of Odontology, Universit

e de Lille,
2
Faculty of Medicine, INSERM U1008, Universit
e de Lille,
3
Department of Prosthodontics, Faculty of Dentistry, Universit

e de Lille,
4
PMOI EA 4490, Universit
e de Lille, Lille,
5
LMCPA, Universit
e de Valenciennes, Valenciennes, France

keywords Abstract
endodontic training; pre-clinical endodontics;
artificial tooth; root canal; hydroxyapatite. Introduction: Endodontic therapy is often complicated and technically demanding.
The aim of this study was to develop a reproducible biomimetic root canal model for
Correspondence pre-clinical and postgraduate endodontic training.
Lieven Robberecht
Faculty of Dentistry
Material and Methods: A specific ceramic shaping technique (3D printing and slip
Universit
e de Lille
casting of a root canal mould) was developed to reproduce canal systems with the
Place de Verdun
desired shape and complexity using a microporous hydroxyapatite (HAp)-based
59000 Lille, France
matrix. The microstructural morphology, pore size and porosity, as well as the Vickers
Tel: +33 320167900
Fax: +33 320167999
microhardness of the ceramic simulators (CS) were assessed and were compared with
e-mail: lieven.robberecht@univ-lille2.fr natural dentin and commercial resin blocks. The reproducibility of the root canal
shapes was assessed using the Dice–Sørensen similarity index. Endodontic treatments,
Accepted: 30 March 2016 from refitting the access cavity to obturation, were performed on the CS. Each step
was controlled by radiography.
doi: 10.1111/eje.12207
Results: Many properties of the CS were similar to those of natural dental roots,
including the mineral component (HAp), porosity (20%, porous CS), pore size
(3.4
2.6 lm) and hardness (120.3
18.4 HV).

Discussion: We showed that it is possible to reproduce the radio-opacity of a tooth

and variations in root canal morphology. The endodontic treatments confirmed that
the CS provided good tactile sensation during instrumentation and displayed suitable
radiological behaviour.

Conclusions: This novel anatomic root canal simulator is well suited for training
undergraduate and postgraduate students in endodontic procedures.

clinical practice (3, 4). According to ESE guidelines, good-qual-

Introduction
Endodontic training constitutes the bulk of dental education
for undergraduate and postgraduate students. Based on Euro-
pean Society of Endodontology (ESE) guidelines, undergraduate
and postgraduate students should be suitably trained to per-
form good-quality root canal treatments (1, 2). Pre-clinical
laboratory exercises are an important step that enables trainees
to acquire the specific manual skills that are essential for
ity endodontic work (GQEW) consists of instrumenting root
canals to an acceptable filling length and performing homoge-
neous root canal fillings (5). A variety of practice models have
been developed to help trainees develop their endodontic skills
or improve their GQEW scores (6, 7).
Extracted human teeth have been used for many years to
teach endodontic procedures (8, 9). However, the disadvantages
of human teeth, including the risk of cross-infection, difficulties

ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd 1
Anatomical root canal simulator
in obtaining them, standardisation and ethical factors, have
stimulated the development of alternative simulators for teach-
ing endodontic procedures.
Simulated root canals in clear epoxy resin blocks for research
and teaching purposes have provided certain benefits. They are
aseptic and have standard lengths, diameters and degrees of
curvature. However, their hardness is totally different from that
of dentin, and certain aspects of the internal anatomy of canals
(3D curvatures, irregularities, natural apical constriction, etc.)
are not accurately reproduced in these blocks (10). In addition,
heat generated inside the canals by instrumentation can soften
the material, deform the preparation and prevent the use of
heated instruments for filling the root canals with gutta-percha
and endodontic sealer. Moreover, their transparency and lack
of radio-opacity restrict the development of some technical
skills related to radiographic interpretation and limit their use
in certain learning phases.
This led to the search for artificial models that reproduce the
anatomical and physical characteristics of dental tissues. For
example, artificial teeth made of opaque resin such as Dentalike
(Dentsply, York, PA, USA) or artificial tooth UMC models
(Smile Factory, Aruja, Brazil) have been reported as alternatives
that better reproduce the features of natural teeth (9). However,
despite the similarities between opaque resin and natural teeth,
these artificial teeth have been limited to studies by endodontic
experts due to substantial differences between the resin and nat-
ural dentin (10).
Hydroxyapatite (HAp) is the main mineral component of
natural teeth. A synthetic analogue has been used to model the
properties of the enamel and dentin of natural teeth (11). The
aim of this work was to develop a reproducible root canal sim-
ulator that consists of an HAp-based root canal system with
the required anatomical features that mimics the chemical
composition of the mineral phase (radio-opacity) and
microstructure (particularly microporosity) of natural dentin. A
specific but simple ceramic shaping technique (3D printing and
slip casting of a root canal mould) was developed (illustrated
in Fig. 1), and the reproducibility of the shape of the root canal
and endodontic treatments using the CS were assessed.
A B
Fig. 1. Fabrication of a root canal simulator by slip casting: (A) root canal mould positioned in a plaster block, (B) hydroxyapatite slurry poured over the
cast and allowed to settle, (C) external view of a sintered artificial root canal simulator.
2 ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Robberecht et al.
Materials and methods
Preparation of the root canal mould
Natural teeth extracted for orthodontic, periodontic or
prosthetic reasons (ethical approval no. DC-2008-642) were
scanned using a micro-CT device (SkyScan 1172; Bruker,
Kontich, Belgium) at 80 kVp, 100 lA and 450 ms integration
time with a voxel size of 20 lm (3). Images of the pulp volume
were reconstructed using NRecon software (Bruker) and were
isolated using CTAn software (Bruker). Micro-CT pulp volume
data were converted into .STL files and were intentionally over-
sized 1.21 times (Meshmixer; Autodesk, San Rafael, CA, USA) to
compensate for shrinkage during HAp sintering. High-resolution
(10 lm) acrylic resin moulds of the canal system were then
3D-printed by stereolithography (CryoCeram; CryoBeryl Software,
Valenciennes, France) using an ultraviolet beam (k = 390 nm,
1 s layer1, and 2.5 mW cm2).
Fabrication of the ceramic root canal system by
slip casting
Hydroxyapatite (HAp) powder was synthesised as previously
described using a classic precipitation approach (12). The cera-
mic root canal system slip casting procedure is illustrated in
Fig. 1. Briefly, the 3D-printed resin canal mould was carefully
placed in a plaster container. A slurry consisting of 75% (dry
matter) hydroxyapatite, 10wt% natural organic porogen (rice
starch; Remy, Wijgmaal, Belgium), 2wt% dispersing agent
(Darvan C; Vanderbilt, Norwalk, CT, USA) and 1.5% binding
agent (Duramax B1001; Rohm and Haas, Philadelphia, PA,
USA) in distilled water was poured over the cast and was
allowed to settle for 3 h.
The green body of the CS was removed from the plaster con-
tainer, allowed to dry, debonded and sintered at 1225°C for 3 h in
a ceramic furnace (LHT 08/17; Nabertherm, Lilienthal, Germany)
to homogeneously densify the HAp and eliminate the porogen
and resin mould. The CS thus obtained had a precise internal root
canal anatomy whilst the external surface was manually shaped.
C
Robberecht et al.
Characterisation of the ceramic root canal
simulator
The CS were sputter coated with gold (HHV auto360; Edwards,
Burgess Hill, UK), and the microstructural morphology and
pore size (n = 10/group) were assessed by scanning electronic
microscopy (SEM, S-3500N; Hitachi, Tokyo, Japan). The porosi-
ties of the CS were measured by water infiltration based on
Archimedes’ method.
A Vickers microindentation hardness test (Testwell, FM) was
performed (ISO 14705:2008 standard) on natural enamel, den-
tin, porous CS (HAp 20), dense CS (HAp) and commercial
plastic blocks (R) (Endo Training Blocks; Dentsply) (n = 10/
group), and the results were compared to the commercial resin
simulator. Briefly, a load (F = 500 g) was applied smoothly to
force the indenter into the test piece. The indenter was held in
place for 10 s. The load was removed, and the impression diag-
onals (d) were measured by SEM. The Vickers hardness value
(HV) was calculated using the following equation:
VH ¼ 0:189  dF2 , where F is the load and d is the impression
diagonal.
The reproducibility of the root canal shape forming process
was assessed by comparing the similarity indexes of eight CS
replicates of the same natural tooth. Briefly, the micro-CT data
of all the CS replicates were acquired by micro-CT scanning.
To embody complex morphological variations, the region of
interest for the analysis was set at 1500 lm from the apex,
within which a series of 20-lm-thick slices were segmented
using CTAn software and imported into open-source 3DSlicer
software by applying a fixed threshold. Dice–Sørensen similarity
coefficients were processed by paired comparison to compare
the morphologies of the different replicates using the DiceCom-
putation module.
Fig. 2. Scanning electronic micrographs (9500) of the microstructures of natural enamel, natural dentin, dense hydroxyapatite (HAp) and porous
hydroxyapatite (HAp 20) produced by adding starch porogen (scale bars = 50 lm).
ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Anatomical root canal simulator
Performing endodontic treatments on the
ceramic root canal simulator
The various root canal treatment steps were performed on the
CS (with a simple or complex root canal system). Access cavi-
ties were refitted with Zekrya-endo and X-Gates (Dentsply)
drills. Root canals were instrumented with manual files (MMC,
MicroMega, Besançon, France) up to a #15 file. The working
lengths were determined on a radiograph with the #15 file
0.5 mm from the apical foramina. The CS were prepared using
the ProTaper system (SX, S1, S2, F1 and F2 files; Dentsply). A
2.5% sodium hypochlorite solution was used to irrigate the
root canal between each instrument change. A .06 tapered
gutta-percha master cone (Dentsply) was then adjusted, and
the root canals were obturated using the thermomechanical
technique. A radiographic assessment was performed before,
during and after endodontic preparation.
Statistical analysis
The statistical difference in microhardness was analysed using
GraphPad Prism 5 (GraphPad Software, La Jolla, CA, USA) by
one-way ANOVA with a Tukey multiple comparisons post-test.
A 95% level of confidence was used.
Results
The CS fabricated by 3D printing and slip casting had a suit-
able macroscopic aspect with no observable fissures. Adding
rice starch as a porogen to the HAp slurry generated a
microstructure (Fig. 2) with a 20% open porosity, which is
much closer to natural dentin (~18% porosity) than acrylic
resin. The pore size in the porous CS was approximately
3
Anatomical root canal simulator Robberecht et al.

A visual comparison (Fig. 4) of the micro-CT images of the

Fig. 3. Vickers microhardness assessment of a natural tooth (enamel,
dentin), a dense HAp ceramic root canal simulator (HAp), a porous HAp
20 ceramic root canal simulator and a commercial plastic block simulator
(R). *signifies significant difference with dentin; **signifies a significant
difference with HAp 20.
4.5
1.3 lm, which was also similar to that of human dentin
(3.4
2.6 lm).
The microhardness test (Fig. 3) showed that the plastic
blocks (R) were softer (37.1
6.82 HV, P < 0.05) than natural
dentin (69
13 HV) whilst the dense CS was harder
(120.3
18.4 HV, P > 0.05). The microhardnesses of natural
dentin and the porous CS (HAp 20) were both significantly
lower than that of enamel and dense HAp (P < 0.05).
canals of the original roots (simple or complex canal system)
and the 3D-printed moulds indicated that the reproductions of
the root canal systems by the CS duplicated the fine details of
the canal system with great fidelity. The average Dice–Sørensen
similarity coefficient was 0.917
0.025, which indicated that
there was a high degree of similarity between the duplicate CS
and, as such, that the fabrication process was highly repro-
ducible.
The endodontic treatments using the CS indicated that it
was feasible to use them to perform the instrumentation
steps. The radiographs (Fig. 4) showed clear canal shapes
(simple or complex) in the CS due to the good radio-opa-
city of the HAp, which facilitated the determination of the
working length, the instrumentation of the root canal system
and the obturation of the root canal (Fig. 4). The CS could
also be used to perform the operations required to access
and refit the cavity (remove dentin triangles and refit the
root canal inlets and isthmus) and to shape and clean the
cavity by mechanical instrumentation and chemical treat-
ments with sodium hypochlorite and EDTA. Heated instru-
ments could also be used for the obturation step (Fig. 5)
without altering the structure of the CS.
Discussion
Endodontic treatments are often complicated and technically
demanding. In vitro models that simulate natural human teeth
are essential for teaching and training dental students in a pre-
clinical setting and dental surgeons in continuing dental educa-
tion courses. Research on endodontic technology also requires

Fig. 4. Design of the shape-controlled reproduction of a (A) simple or (B) complex artificial root canal system in a ceramic simulator. From left to right:
reconstructed micro-CT image of a pulp volume, a 3D-printed mould of a pulp volume, a pre-operative radiograph of a root canal simulator, a
radiograph of a gutta-percha master cone in the root canal simulator and a radiograph of a well-filled root canal simulator.

4 ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Robberecht et al.
A B
C D
Fig. 5. Endodontic treatment of a complex (mandibular molar) root canal simulator. (A) pre-operative view, (B) pulp chamber roof removed, (C) canal
inlets and mesial isthmus refitted, (D) endodontic filling with warm gutta-percha compaction and sealer.
good models for in vitro performance assessments prior to
using the technology on patients. The drawbacks of existing
models (extracted human teeth, animal teeth and simulated
root canals in epoxy resin blocks, for example) and the lack of
suitable alternatives make the development of new models for
educational and research purposes a priority (10).
The biomimetic microporous HAp-based dental root canal
model was fabricated by 3D resin canal mould printing
(stereolithographic technique) and simple slip casting. Whilst
the fabrication technique is complex, it can be easily ramped
up to an industrial scale, which would reduce the cost per unit.
This novel CS may potentially better meet the specific require-
ments (radio-opacity, hardness, shape) of endodontic teaching
and training.
Dentin is a natural material whose porosity, mineralisation
and microstructure vary with the anatomy, age and pathologi-
cal history of the teeth (11). These variations limit the repro-
ducibility of assays based on natural teeth in certain contexts.
The addition of rice starch (porogen) to the hydroxyapatite
slurry produced a homogeneous microporous structure. The
porosity of ceramic matrixes can be tailored to meet specific
needs such as simulating young or sclerotic dentin by varying
the amount of porogen added to the slurry. The rice starch was
milled and was mixed with the HAp powder to ensure that the
porogen was evenly distributed throughout the HAp slurry.
The sintering step burned out the porogen, giving the CS a
microporous structure. Further research is required to fabricate
CS in which the pores around the pulp cavity are oriented in
an annular fashion (13).
ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Anatomical root canal simulator
Morphological variations of the root canal system are com-
mon difficulties encountered in endodontic treatments and also
present a major challenge when reproducing root canal systems
in simulators. The high precision of the 3D printing technique
(10 lm) makes it possible to reproduce any canal anatomy
from micro-CT data, and the slip casting can be easily adapted
to fabricate pulpal moulds of any complexity (from single to
multiple canals) as shown in Fig. 4. The flexibility of the CS
fabrication process thus shows great promise for ramping up to
an industrial scale.
Transparent epoxy root canal simulators have been used for
many years as many clinicians consider that they provide the
visibility required to teach endodontic techniques. LaTurno
et al. (14) showed that epoxy resin simulators based on trans-
parent blocks make it easier to study and understand the
biomechanics of root canal shaping. Whilst the CS described
here clearly does not have this ‘advantage’, it does provide a
model that closely recapitulates a clinical setting and that
requires dental students and clinicians to rise to the challenge
posed by potential anatomical difficulties and pitfalls and to
perfect their technique (2, 4). As the fabrication of CS is a
complex process, we have not yet reproduced the crown shape.
Whilst the absence of the crown shape is a drawback for daily
use in pre-clinical education, future research will concentrate
on resolving this limitation.
Epoxy resin simulators are radio-translucent. As such, the
results of treatments cannot be assessed using radiographs with-
out applying a coating of bismuth oxide varnish (15). However,
the radio-opacity and contrast of HAp is similar to that of
5
Anatomical root canal simulator
natural dentin, providing good-quality radiographic images
of endodontic materials. The use of resin is also limited due to
insufficient hardness as well as softening during instrumentation,
which is not the case for HAp. As shown in Fig. 3, the CS
is harder (120.3
18.4 HV) than a natural root canal
(69
13 HV), which is critical for tactile sensation during
instrumentation. This was confirmed by endodontic treatments
performed on the CS.
The process described here makes it possible to reproducibly
fabricate any shape found in endodontic classifications by addi-
tive stereolithography. It can also be used to adjust the diffi-
culty of training exercises by reproducing ledge formations,
root perforations and blockages. The difficulty of training exer-
cises and student evaluations can thus be standardised. Lastly,
like resin reproductions of natural teeth (16), our CS also
allows dentists to practise on a reproduction of a tooth with a
complex root configuration before beginning the actual
endodontic treatment. Cone beam-computed tomography data
can also be collected before surgery, decreasing the risk of
iatrogenic accidents and facilitating the prognosis of the
endodontic treatment.
Conclusion
The new CS described here reproduces a number of important
features (composition, microstructure, hardness, canal anat-
omy) that are much closer to those of natural teeth than those
of simple plastic blocks. The fabrication process can be easily
adapted to produce any desired canal shape (simple or
complex) as well as specific traits of dentin. It is also highly
reproducible, making it possible to manufacture duplicates in
large quantities. This novel CS also poses no risk of infection
and makes objective assessments possible due to its uniformity.
It is thus well suited for practising endodontic procedures, can
be mounted on phantom heads and can be used to determine
the working length with an apex locator. It is also suitable for
training students and conducting research on new endodontic
technologies. Further research is required to fabricate a truly
representative CS in which the external form, especially the
anatomic crown, is associated with the anatomy of the internal
pulp and in which the pores around the pulp cavity are
oriented in an annular fashion.
Acknowledgements
The authors thank Vincent Hornez (CryoBeryl Software) for
supplying the 3D-printed root canal systems.
L. Robberecht, F. Chai, JC. Hornez, M. Dehurtevent, T.
B
ecavin and E. Deveaux have a patent: Simulateur canalaire
endodontique artificiel a base d’hydroxyapatite. BFF140597DB
6 ª 2016 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Robberecht et al.
BCDG. P. Marchandise declares no potential conflict of inter-
est.
References
1 De Moor R, H€ ulsmann M, Kirkevang LL, Tanalp J, Withworth J.
Undergraduate curriculum guidelines for endodontology. Int Endod
J 2013: 46: 1105–1114.
2 European Society of Endodontology. Accreditation of postgraduate
specialty training programmes in endodontology. Minimum criteria
for training specialists in endodontology within Europe. Int Endod
J 2010: 43: 725–737.
3 European Society of Endodontology. Quality guidelines for
endodontic treatment: consensus report of the European Society of
Endodontology. Int Endod J 2006: 39: 921–930.
4 Sonntag D, B€arwald R, H€ ulsmann M, Stachniss V. Pre-clinical
endodontics: a survey amongst German dental schools. Int Endod J
2008: 41: 863–868.
5 Tchorz JP, Hellwig E, Altenburger MJ. Teaching model for
artificial teeth and endodontic apex locators. J Dent Educ 2013:
77: 626–629.
6 Pitts NB, Chow T, Renson CE, Kirkwood J, Walker RT. Mounted
human skulls as an aid to preclinical endodontic teaching. Int
Endod J 1984: 17: 73–75.
7 Dummer PMH, Alodeh MHA, Al-Omari MAO. A method for the
construction of simulated root canals in clear resin blocks. Int
Endod J 1991: 24: 63–66.
8 Spenst A, Kahn H. The use of a plastic block for teaching root
canal instrumentation and obturation. J Endod 1979: 5: 282–284.
9 Nassri MRG, Carlik J, da Silva CRN, Okagawa RE, Lin S. Critical
analysis of artificial teeth for endodontic teaching. J Appl Oral Sci
2008: 16: 43–49.
10 Tanalp J, G€ ung€
or T. Apical extrusion of debris: a literature review
of an inherent occurrence during root canal treatment. Int Endod J
2014: 47: 211–221.
11 Pashley DH, Andringa HJ, Derkson GD, Derkson ME, Kalathoor
SR. Regional variability in the permeability of human dentin. Arch
Oral Biol 1987: 32: 519–523.
12 Hornez JC, Chai F, Monchau F, Blanchemain N, Descamps M,
Hildebrand HF. Biological and physic-chemical assessment of
hydroxyapatite (HA) with different porosity. Biomol Eng 2007: 24:
505–509.
13 Chevalier E, Chulia D, Pouget C, Viana M. Fabrication of porous
substrates: a review of processes using pore forming agents in the
biomaterial field. J Pharm Sci 2008: 97: 1135–1154.
14 LaTurno SAL, Corcoran JF, Ellison RL. An evaluation of a teaching
aid in endodontics. J Endod 1984: 10: 507–511.
15 Tchorz JP, Brandl M, Ganter PA, et al. Pre-clinical endodontic
training with artificial instead of extracted human teeth: does the
type of exercise have an influence on clinical endodontic outcomes?
Int Endod J 2015: 48: 888–893.
16 Byun C, Kim C, Cho S, et al. Endodontic treatment of an
anomalous anterior tooth with the aid of a 3-dimensional printed
physical tooth model. J Endod 2015: 41: 961–965.