doi:10.1111/iej.

13141

Anatomical danger zone reconsidered: a micro-CT

study on dentine thickness in mandibular molars

G. De-Deus1 , E. A. Rodrigues1, F. G. Belladonna1 , M. Simo ~ es-Carvalho1 ,

1 1 2
D. M. Cavalcante , D. S. Oliveira , E. M. Souza , K. A. Giorgi , M. A. Versiani1
1
,
3 1,4
R. T. Lopes , E. J. N. L. Silva & S. Paciornik5
1
Department of Endodontics, Fluminense Federal University, Niter oi, Rio de Janeiro; 2Department of Dentistry II, Federal
3
University of Maranh~ ao, S~ ao; Nuclear Engineering Program, Federal University of Rio de Janeiro, Rio de
ao Luıs, Maranh~
Janeiro, Rio de Janeiro; 4Department of Endodontics, Grande Rio University, Duque de Caxias, Rio de Janeiro; and 5Department
of Chemical and Materials Engineering, Pontifıcia Universidade Cat
olica do Rio de Janeiro, Rio de Janeiro, Rio de Janeiro, Brazil

Abstract
De-Deus G, Rodrigues EA, Belladonna FG, Simo ~ es-
Carvalho M, Cavalcante DM, Oliveira DS, Souza
EM, Giorgi KA, Versiani MA, Lopes RT, Silva EJNL,
Paciornik S. Anatomical danger zone reconsidered: a micro-
CT study on dentine thickness in mandibular
International Endodontic Journal.
Aim To investigate the smallest dentine thickness in
mesial canals of mandibular molars along the cervical
and middle thirds of the root by means of a micro-
computed tomographic (micro-CT) technology and
digital image analysis.
Methodology Fifty mesial roots of mandibular
molars having two independent canals (mesiobuccal
and mesiolingual), in the coronal and middle levels,
were selected and scanned in a micro-CT device. After
reconstruction procedures, approximately 468 slices
per root covering the 7 mm below the furcation area
of the mesial root were analysed to measure the
smallest dentine thickness (danger zone [DZ]) in each
slice from both distal and mesial regions of the mesial
canals by an automatic segmentation process.
Results The DZ values in the mesiobuccal canals
varied from 0.67 to 1.93 mm, with an average of
1.13  0.21 mm. For the mesiolingual canals, the
DZ varied from 0.77 to 1.89 mm with an average of
molars.
1.10  0.21 mm. There was no correspondence in
the DZ between the mesiobuccal and mesiolingual
canals at the same cross-sectional level in 71% of the
specimens. Moreover, the smallest dentine thickness
was towards the mesial region of the roots in 22%
and 18% of the mesiolingual and mesiobuccal canals,
respectively.
Conclusions The smallest dentine thickness was
on the mesial plane of the roots in about 40% of the
canals. The vertical location of the DZ in relation to
the furcation area was in the middle third of the root.
Keywords: danger zone, dentine thickness,
mandibular molars, mesial roots, micro-CT.
Received 21 March 2019; revision requested 2 May 2019,
accepted 3 May 2019

canal treatment (Estrela et al. 2018). These perfora-

Introduction
Midroot perforations are normally caused by over-in-
strumentation on an already thin dentinal wall,
which may seriously compromise the outcome of root
Correspondence: Gustavo De-Deus, Av. Henrique Dodsworth
85 Apto 808 - Lagoa, Rio de Janeiro, RJ, zip code: 22061-
030, Brazil (e-mail: endogus@gmail.com).
tions have been related historically to the distal area
of mesial roots in mandibular molars, and thus,
Abou-Rass et al. (1980) introduced the concept of the
‘danger zone’ (DZ) in the early 1980s. In fact, these
authors formally reported what experienced clinicians
already knew: often, mesial canals of mandibular
molars do not assume a central position in the root
with the distal area between the canal and root bifur-
cation being relatively thin, the so-called DZ, which is

© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd International Endodontic Journal 1
 Anatomical danger zone reconsidered De-Deus et al.
more vulnerable to strip perforations. On the other
hand, the safety zone was described as the mesial
area of the mesial root with a thicker dentine layer,
which is often minimally instrumented by endodontic
instruments. In short, Abou-Rass et al. (1980) pointed
out the importance of this anatomical area during
canal shaping and, since then, many studies have
evaluated the safety of various preparation techniques
in mesial canals of mandibular molars (Garcia Filho
et al. 2003, Akhlaghi et al. 2010, Silva et al. 2017).
The anatomy of the DZ as well as the assessment of
canal preparation techniques and instruments has
been based on destructive and invasive approaches
(i.e. sectioning methods) (Garcia Filho et al. 2003,
Akhlaghi et al. 2010). This leads to serious experi-
mental limitations as it allows the analysis of only a
few slices per root.
In the early 2000s, the introduction of micro-com-
puted tomographic (micro-CT) imaging opened new
possibilities for endodontic research since this technol-
ogy allows precise nondestructive longitudinal two-di-
mensional (2D) and three-dimensional (3D)
assessments (Peters et al. 2001, 2003, De-Deus et al.
2015, 2016, Silva et al. 2017), because micro-CT is
based on a powerful X-ray source, which allows visu-
alization and measurement of the outer and inner
structures of an opaque object without needing previ-
ous sample preparation or chemical fixation. For
instance, micro-CT with mathematical modelling has
provided information on the dentine thicknesses in 1-
mm intervals (Harris et al. 2013). However, detailed
morphological information about the DZ remains
inconsistent, sparse and sometimes controversial, as
the evaluation of just a few cross sections per sample
is rather meaningless, especially when using micro-
CT technology. Lee et al. (2015) appears to be the
only micro-CT study that used high-resolution and
smaller slice intervals (0.1 mm). Therefore, a compre-
hensive anatomical study of danger and safety zones
in mesial roots of mandibular molars is timely and
may help reduce either the risk of midroot perfora-
tions towards the furcation area or an unnecessary
loss of sound dentine tissue, which appears to con-
tribute to the long-term survival of teeth (Soares et al.
2008).
The current study has a descriptive nature and was
designed to investigate the smallest dentine thickness
in mesial canals of mandibular molars along the cer-
vical and middle thirds of the root by means of micro-
CT technology and digital image analysis.
2 International Endodontic Journal
Materials and methods
Sample selection and imaging
The present ex vivo study was approved by the Ethical
Research Committee of Fluminense Federal University.
Initially, one hundred moderately curved mesial roots
(10–20°) of first and second mandibular molars,
selected according to Schneider’s method (Schneider
1971), with 10  1 mm in length, were imaged with
a micro-CT scanner (SkyScan 1173; Bruker microCT,
Kontich, Belgium) at 14.25 lm (pixel size), 70 kV,
114 mA, 180° rotation around the vertical axis, rota-
tion step of 0.7°, camera exposure time of 250 mil-
liseconds and frame average of 4, using a 1-mm-thick
aluminium filter. The images were reconstructed
(NRecon v. 1.7.1.6; Bruker microCT) with similar
parameters for beam hardening (35–45%), ring arte-
fact correction (3–5) and contrast limits (0–0.05).
Then, fifty mesial roots with two independent canals
(mesiobuccal and mesiolingual), at the coronal and
middle levels, were selected. None of the specimens
had a root filling, root caries, cracks, fractures, and
internal or external resorption.
Image analysis
The volume of interest was selected extending from
the furcation level to 3 mm from the anatomical
apex of the mesial roots, corresponding to approxi-
mately 468 slices per root, for a total of 23 400
slices. An image analysis routine was developed to
measure the minimal dentine thickness from both
distal and mesial aspects of the mesial canals of
mandibular molars by using a previously validated
BoneJ plug-in (Doube et al. 2010) implemented into
the Fiji/ImageJ software (Fiji v.1.51n; Madison, WI,
USA). First, a 3D median filter was applied to the
stacks in order to reduce overall noise (Neves et al.
2015) and the dentine was binarized using a mini-
mum threshold algorithm. Then, the BoneJ plug-in
was used to identify the position and measure the
smallest dentine thickness in each slice for both
mesial canals.
Three-dimensional mapping of the dentine thick-
ness was created and saved for structure thickness in
CTAn v.1.15 software (Bruker microCT), and loaded
into CTVox v.3.3 software (Bruker microCT) to gener-
ate colour-coded 3D models of the mesial roots of
mandibular molars.
© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd
 De-Deus et al. Anatomical danger zone reconsidered

Table 1 Descriptive data regarding the smallest dentine

thickness (DZ) and location of the DZ Results

Thickness of Thickness of Location of Table 1 presents descriptive data for all specimens
the DZ in the DZ in the DZ under regarding the smallest dentine thickness (DZ) and its
mesiobuccal mesiolingual the furcation location relative to the furcation area. Table 2 shows
canals (mm) canals (mm) area (mm) the location of the DZ for all specimens according to
Mean  1.13  0.22 1.10  0.21 4.37  1.68 the distance from the furcation area distributed in 1-
SD mm intervals.
Range 0.67–1.93 0.77–1.89 1.1–7
The DZ values in the mesiobuccal canals varied from
0.67 to 1.93 mm, with an average of
Table 2 Distribution of the DZ location along the cross sec- 1.13  0.21 mm. For the mesiolingual canals, the DZ
tions for all specimens according to the distance from the varied from 0.77 to 1.89 mm, with an average of
furcation area
1.10  0.21 mm. There was no correspondence in the
Location of the DZ below the % of cross Cumulative DZ between the mesiobuccal and mesiolingual canals
furcation area sections % at the same cross-sectional level in 71% of the speci-
1–2 mm 9.7 9.7 mens. Moreover, the DZ was towards the mesial region
2–3 mm 13 22.7 of the roots in 22% and 18% of the mesiolingual and
3–4 mm 12.8 35.5 mesiobuccal canals, respectively (Figs 1 and 2).
4–5 mm 27.1 62.6
Figure 3 shows colour-coded representations of the
5–6 mm 21 83.6
6–7 mm 16.4 100
dentine thickness throughout mesial roots of five rep-
resentative mandibular molars. Qualitative analysis

Figure 1 Cross-sectional images of 3 mesial roots of mandibular molars showing the noncentred position of the mesial canals.
a1, a2 and a3 show the thinnest dentine towards the mesial area in the mesiolingual canal and the distal area in the
mesiobuccal canal. b1, b2 and b3 show the DZs towards the distal area in both mesiolingual and mesiobuccal canals. c1, c2
and c3 show the DZs towards the mesial area in both mesiolingual and mesiobuccal canals.

© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd International Endodontic Journal 3
 Anatomical danger zone reconsidered De-Deus et al.

demonstrated that the noncentred position of the
mesial canals and the asymmetric shape of the root
resulted in variable thickness of the dentine at differ-
ent levels and directions, including areas towards the
mesial aspect of the root in some cases.
Discussion
The current study provided two innovative results
regarding the anatomical DZ. First, the smallest den-
tine thickness was towards the mesial plane of the
roots in around 40% of the canals (22% and 18% of
the mesiolingual and mesiobuccal canals, respec-
tively). Secondly, the vertical location of the DZ in
relation to the furcation area was towards the middle
third of the root.
Relevant original data revealed by the present
study were about the position of the DZ in the cross-
sectional plane of the root. Traditionally, the under-
standing from the classic concept of the DZ refers to
the distal area between the main canal space and the
root bifurcation, which has the thinnest dentine and
is more prone to the development of strip perforations.
However, the current results revealed that the DZ
was displaced towards the furcation area in only 60%
of the evaluated cross sections. In the other 40% of
the slices, the thinnest dentine was located on the
mesial region of the roots, which contrasts with the
generally held view (Abou-Rass et al. 1980). Also
using a micro-CT imaging analysis, Lee et al. (2015)
found the thinnest root canal wall on the mesial por-
tion of the root in between 15% and 33% of the spec-
imens, which is consistent with the present results.
The DZ evaluated in the present study was located
up to 4 mm under the furcation area in only 35% of
the specimens, whilst the majority of the samples
revealed that the DZ was between 4 and 7 mm below
the furcation area. In other words, DZs prone to
either strip perforation or unnecessary loss of dentine
are more towards the middle third of the root
(4.37  1.68 mm under the furcation area) than pre-
viously reported (Table 1). This finding is innovative

Figure 2 Cross-sectional images of 3 mesial roots of mandibular molars showing the noncentred position of the mesial canals.
d1, d2 and d3 show the thinnest dentine towards the distal area in the mesiolingual canal and the mesial area in the
mesiobuccal canal. e1, e2 and e3 show the DZs towards the distal area in both mesiolingual and mesiobuccal canals. f1, f2
and f3 show the DZs towards the mesial area in both mesiolingual and mesiobuccal canals.

4 International Endodontic Journal © 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd
 De-Deus et al. Anatomical danger zone reconsidered

Figure 3 Three-dimensional colour-coded models of five mandibular molars showing that noncentred position of the mesial
canals and the asymmetric shape of the root resulted in variable dentine thickness at different levels and directions of the
roots.

© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd International Endodontic Journal 5
 Anatomical danger zone reconsidered De-Deus et al.
inasmuch as previous information described the
anatomical position of the DZ is concentrated up to
4 mm below the furcation level (Kessler et al. 1983,
Berutti & Fedon 1992, Garcia Filho et al. 2003,
Sauaia et al. 2010, Tabrizizadeh et al. 2010, Akhlaghi
et al. 2015).
It is worth mentioning that the mean smallest den-
tine thickness along the distal surfaces in mesial roots
of mandibular molars found in this study was
0.67 mm, which is smaller than usually reported in
the literature: Lim & Stock (1987) = 0.94 mm, Garcia
Filho et al. (2003) = 0.79 mm, Kessler et al. (1983) =
1.08 mm, Akhlaghi et al. (2015) = 1.05 mm, Berutti
& Fedon (1992) = 1.2 mm and Tabrizizadeh et al.
(2010) = 1.3 mm. The information on the smallest
dentine thickness values was mainly created from
investigations based on destructive sectioning meth-
ods and direct optical microscopy observation of a few
slices of the mesial roots. One exception is a cone
beam computed tomography study with low spatial
resolution (Akhlaghi et al. 2015). The other exception
is a micro-CT study which revealed values of
0.81 mm, but the measurements were taken only
1.5 mm below the furcation area (Harris et al. 2013).
The original findings presented in this study are
probably a consequence of the methodological
advancement achieved with the interplay between the
micro-CT imaging technology and an automatic com-
putational routine for digital image analysis and pro-
cessing. In fact, the accumulated body of evidence on
the DZ was created essentially based on destructive
methods and direct microscopy observation of a few
root sections per tooth (Kessler et al. 1983, Berutti &
Fedon 1992, Garcia Filho et al. 2003, Sau aia et al.
2010, Tabrizizadeh et al. 2010). Moreover, even stud-
ies that used micro-CT imaging technology only per-
formed the evaluation of a few cross-sectional images
(Harris et al. 2013, Ordinola-Zapata et al. 2019). In
this study, a full 3D mapping of dentine thickness
along the entire cervical and middle thirds was per-
formed rendering data from hundreds of cross sections
per root. Furthermore, measurements performed by
an automatic computational routine are more robust
and reliable. Markedly, the automated analysis
allowed the rapid measurement of thousands of slices
rendering the experiment less time-consuming and
labour-intensive. Moreover, the age of teeth is an
uncontrolled variable that may have influenced the
present results at least to some degree. Stored teeth
were used with unknown information about age; as
6 International Endodontic Journal
long as age affects the overall canal size, it is likely to
affect the overall dentinal thickness.
Conclusions
The smallest dentine thickness in extracted mandibu-
lar first and second molars was towards the mesial
plane of the roots in approximately 40% of the
canals, whilst the overall vertical location of the DZ
was towards the middle third of the root.
Conflict of interest
The authors have stated explicitly that there are no
conflicts of interest in connection with this article.
References
Abou-Rass M, Frank AL, Glick DH (1980) The anticurvature
filing method to prepare the curved root canal. Journal of
the American Dental Association 101, 792–4.
Akhlaghi NM, Kahali R, Abtahi A, Tabatabaee S, Mehrvarz-
far P, Parirokh M (2010) Comparison of dentine removal
using V-taper and K-Flexofile instruments. International
Endodontic Journal 43, 1029–36.
Akhlaghi NM, Bajgiran LM, Naghdi A, Behrooz E, Khalilak Z
(2015) The minimum residual root thickness after using
ProTaper, RaCe and Gates-Glidden drills: a cone beam
computerized tomography study. European Journal of Den-
tistry 9, 228–33.
Berutti E, Fedon G (1992) Thickness of cementum/dentin in
mesial roots of mandibular first molars. Journal of Endodon-
tics 18, 545–8.
De-Deus G, Belladonna FG, Silva EJ et al. (2015) Micro-CT
evaluation of non-instrumented canal areas with different
enlargements performed by NiTi systems. Brazilian Dental
Journal 26, 624–9.
De-Deus G, Belladonna FG, Marins JR et al. (2016) On the
causality between dentinal defects and root canal prepara-
tion: a micro-CT assessment. Brazilian Dental Journal 27,
664–9.
Doube M, Kłosowski MM, Arganda-Carreras I et al. (2010)
BoneJ: free and extensible bone image analysis in ImageJ.
Bone 47, 1076–9.
Estrela C, Decurcio DA, Rossi-Fedele G, Silva JA, Guedes OA,
Borges AH (2018) Root perforations: a review of diagnosis,
prognosis and materials. Brazilian Oral Research 32, e73.
Garcia Filho PF, Letra A, Menezes R, Carmo AM (2003)
Danger zone in mandibular molars before instrumenta-
tion: an in vitro study. Journal of Applied Oral Sciences 11,
324–6.
Harris SP, Bowles WR, Fok A, McClanahan SB (2013) An
anatomic investigation of the mandibular first molar using
© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd

micro-computed tomography. Journal of Endodontics 39,
1374–8.
Kessler JR, Peters DD, Lorton L (1983) Comparison of the
relative risk of molar root perforations using various
endodontic instrumentation techniques. Journal of Endodon-
tics 9, 439–47.
Lee JK, Yoo YJ, Perinpanayagam H et al. (2015) Three-di-
mensional modelling and concurrent measurements of
root anatomy in mandibular first molar mesial roots using
micro-computed tomography. International Endodontic Jour-
nal 48, 380–9.
Lim SS, Stock CJ (1987) The risk of perforation in the
curved canal: anticurvature filing compared with the
stepback technique. International Endodontic Journal 20,
33–9.
Neves AA, Silva EJ, Roter JM et al. (2015) Exploiting the
potential of free software to evaluate root canal biome-
chanical preparation outcomes through micro-CT images.
International Endodontic Journal 48, 1033–42.
Peters OA, Sch€ onenberger K, Laib A (2001) Effects of four
Ni-Ti preparation techniques on root canal geometry
assessed by micro computed tomography. International
Endodontic Journal 34, 221–30.
Ordinola-Zapata R, Martins JNR, Versiani MA, Bramante CM
(2019) Micro-CT analysis of danger zone thickness in the
© 2019 International Endodontic Journal. Published by John Wiley & Sons Ltd
De-Deus et al. Anatomical danger zone reconsidered
mesiobuccal roots of maxillary first molars. International
Endodontic Journal 52, 524–9.
Peters OA, Peters CI, Sch€ onenberger K, Barbakow F (2003)
ProTaper rotary root canal preparation: effects of canal
anatomy on final shape analysed by micro CT. Interna-
tional Endodontic Journal 36, 86–92.
Sau aia TS, Gomes BP, Pinheiro ET et al. (2010) Thickness of
dentine in mesial roots of mandibular molars with differ-
ent lengths. International Endodontic Journal 43, 555–9.
Schneider SW (1971) A comparison of canal preparations in
straight and curved root canals. Oral Surgery Oral Medicine
Oral Pathology 32, 271–5.
Silva EJNL, Pacheco PT, Pires F, Belladonna FG, De-Deus G
(2017) Microcomputed tomographic evaluation of canal
transportation and centring ability of ProTaper Next and
Twisted File Adaptive systems. International Endodontic
Journal 50, 694–9.
Soares PV, Santos-Filho PC, Queiroz EC et al. (2008) Frac-
ture resistance and stress distribution in endodontically
treated maxillary premolars restored with composite resin.
Journal of Prosthodontics 17, 114–9.
Tabrizizadeh M, Reuben J, Khalesi M, Mousavinasab M, Eza-
badi MG (2010) Evaluation of radicular dentin thickness
of danger zone in mandibular first molars. Journal of Den-
tistry (Tehran) 7, 196–9.
International Endodontic Journal 7