Micro CT settings for caries detection: how to optimize

Pereira-Oliveira S et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 5, Issue 11, (Part - 5) November 2015, pp.78-82
www.ijera.com 78 | P a g e
Micro CT settings for caries detection: how to optimize.
Pereira-Oliveira S*; Chaiben CL**; Ando M***; Freitas CF****; Zero D***;
Azevedo-Alanis LR**; Ferreira-Zandona AG***.
* MD, PhD, Professor at ABO - Brazilian Dental Association.
**School of Health and Biosciences, Pontifícia Universidade Católicado Paraná (PUCPR), Curitiba, Brazil.
*** Oral Health Institute, Indiana University 415 Lansing Street Indianapolis, IN 46202-2876
**** University of São Paulo Av. Professor Lineu Prestes, 2227 Cidade Universitária– São Paulo – SP
ABSTRACT
Some important items that can influence micro CT image were reviewed in this study. Different settings were
optimized for the assessment of early caries lesions. There are several researches on bone using micro CT but not
too much on dental hard tissues when assessing mineral loss. Different kinds of micro CT devices and
technologies are taking place today, each requiring unique settings, and this consists one of the greatest obstacles
for the use of micro CT on dental hard tissues. Achieving the settings for an ideal dental image is therefore a
challenge. The purpose of this study was to evaluate different micro CT settings to optimize the assessment of
early caries lesions aiming the integrity of the dental specimen thus, making possible to reuse it for further
studies. Three teeth with early caries lesions were submitted to different micro CT settings and different
reconstruction settings, aiming a better image. The final image was compared visually through different densities
and attenuation coefficients. The best setting for teeth tissues was achieved regarding contrast, definition, noise
reduction and the larger difference between sound enamel and early lesions attenuation coefficient.
Keywords - Dental Caries, Diagnostic Imaging, Radiographic Image Enhancement, X-Ray Microtomography
I. INTRODUCTION
The ability of micro CT to show bone
morphology and microstructure and provide
quantitative mineralization assessment of calcified
tissue1
makes it an important tool to assess hard
tissues. It provides fast, non-destructive assessment
of bone architecture2
, allowing the evaluation of
spatial differences of specimens, due to several
digital sections that it can provide3
. It has been
important to increase our understanding of some
disease progression and the efficacy of therapies to
arrest or slow hard tissue loss. The development of
3D images and of analysis software makes it possible
to have qualitative measures that can be extracted
from computer data3
.
Due to the many characteristics of the micro CT
there has been interest in this technique for
evaluation of dental hard tissues. Some of the
challenges presented are the number of micro CT
apparatus available, each operating with different
settings. Some micro CT have the same technology
as the cone beam scanner4
and this apparatus is being
improved through the years. As it is being used in
many different researches, achieving the settings for
a feasible image is still a challenge. There have been
several studies on bone using microCT5,6,7,8,9,10
but
not very much on dental hard tissues11,12,13,14,15
.
Phantoms are objects with known composition
and density, used to calibrate CT settings for bone
assessment12,16
. Phantoms might be used to correct
sensors non-uniformities, ensuring sensor optimal
working condition12
, reduction of beam hardening
and to provide a good approximation of equivalent
monochromatic X-ray attenuation for biological
specimens13
. A suitable micro CT reference material
to calibrate the scanner for hard tissue measurements
might mimic the absorption properties of the
material. Increasing mass attenuation coefficients
corresponding to an exactly determined concentration
of bone mineral have to be attained. Also, the
reference material has to be homogenous on a
micrometer scale such that the standard deviation of
the attenuation coefficient in a given area is reduced
to a minimum16
. But even with phantoms, the
methods for calibration and accuracy have not been
thoroughly evaluated20
.
Several marketed phantoms are made of
hydroxyapatite (HA), most of them, with HA ashes
compressed through a cylinder. Other studies use a
step wedge of aluminum or other aluminum object
(as a small cylinder)19
. All phantoms have limitations
as short-term reproducibility of phantoms
measurements, and accuracy of bone tissue mineral
densities measurements by correlation to ash
density20
. This affirmative increases the difficulty in
determining optimal settings, considering the
differences in size, macro and micro architecture,
between different hard tissues, even from the same
patient.
RESEARCH ARTICLE OPEN ACCESS

Most of marketed phantoms are made
considering the bone attenuation coefficient, and their
use for dental hard tissues is limited as the enamel
attenuation coefficient is extremely higher when
comparing with bone.As there are no marketed
phantoms to be used when assessing enamel, several
considerations must be made prior to decide theideal
settings for evaluation of dental tissues.
Of the first consideration when imaging a tissue
is the production of artifacts. Artifacts are any
systematic discrepancy between the CT numbers in
the reconstructed image and the true attenuation
coefficient of the object3
. In computed tomography,
artifacts can alter the image more often than
conventional radiographs due to a great number of
detector measurements. When acquiring microCT
images two possible artifacts can occur: physics-
based artifact and scanner-based artifacts3
.
Physics-based artifacts results from the
acquisition data. The most common physics-based
artifact is beam hardening that can produce cupping
artifacts and streaks and dark bands. To minimize
beam hardening filtration, calibration correction and
beam hardening correction software can be used.
Filtration is used to make the beam harder before
reaching the object. It has the objective to eliminate
the lower-energy photons before it passes through the
specimen. With a more uniform beam, the X-ray
photons will not reach the detectors more intensively,
this way, avoiding the described artifacts. If a good
filtration is not performed, a multi-energy beam can
reach the specimen and by the time it is passing
through the object, the lower photons are absorbed,
making the beam harder17
.
Calibration Correction is frequently made with
phantoms in a range of sizes to allow the detectors to
be calibrated with compensation for the beam
hardening effects of different tissue of the specimen.
However, there are no marketed phantoms designed
for dental tissues and natural tissues never exactly
match with the phantom’sshape. Enamelis the hardest
issue on human body and its thickness must also be
considered for further analysis. It is also important to
perform flat field optimization to correct the images
recorded by the detectors, not only with the phantoms
but with each specimen scanned. Flat field
optimization helps to calibrate the detectors avoiding
beam hardening effects17
.
Beam hardening correction software is based on
algorithms applied to the images during
reconstruction.
The number of projections used to reconstruct a
microCT image is one of the determining factors in
image quality. The undersamplingmode decreases the
reconstruction time but may create a too large
interval between projections.Undersamplingcan
produce an effect known as view aliasing due to
misregistration by the computer of information
relating to sharp edges. This effect can create fine
stripes distant from a dense structure17
.
Scanner based artifacts results from
imperfections in scanner function. This malfunction
may cause ring artifacts, which are one of the most
common problems with tomographic images. Ring
artifacts result from detectors out of calibration. In
high-resolution microCT using flat detectors (FD),
imperfect or defect detector elements may cause
concentric-ring artifacts due to their continuous over-
or underestimation of attenuation values, which often
disturb image quality18
. The noncalibrated detector
will read erroneously each angular position, resulting
in a circular artifact. Even when these artifacts are
visible they will rarely be confused with a lesion.
However, to improve image quality, flat field
correction can be performed during the acquisition
and the reconstruction software has the option of ring
artifact reduction that can be used at its maximum
performance.
Dental caries remains one of the most prevalent
diseases. The use of fluoride has changed its
behaviorslowing its progression and bringing
challenges to its detection. Early detection of caries
lesions is increasingly more significant as
management of dental caries at an early stage
becomes the norm. New tools for early caries
detection are even more investigated, such as
microCT, and validation of this methodology also
becomes crucial.(3-4)19,20
Histology has been used as a validation method
on many in vitro dental caries studies.Although
capable of detecting minute changes, it requires
destruction of the specimen, limiting its use when
longitudinal assessment of specimens would be
desirable.MicroCT produces images with excellent
resolution and is being used in dentistry to evaluate
the mineral content of hard tissues due to its accuracy
in measuring the linear attenuation coefficient.21,22
The purpose of this study is to systematically
evaluate different microCT settings on the
assessment of early caries lesions.
II. Material and Methods
Three teeth were collected in the premises of
Indiana University School of Dentistry (IUSD,
Indianapolis IN, USA) (IRB #0306-64 exempt).
Teeth were kept in a solution of Thymol 0,1%
between the scans. They were cleaned with a
Robinson brush and water and, the coronal surface
was evaluated using the ICDAS criteria22,23,24,25
by
one trained examiner. As the goal is to optimize the
settings for assessment of early lesions the3teeth
selected (1 molar and 2 premolars) had ICDAS codes
0-2 on proximal surfaces:
Specimen 1: ICDAS code 0 on both surfaces;

Specimen 2: ICDAS code 0 on one proximal surfaces
and 1 on the other;Specimen 3: ICDAS code 2 on
both surfaces.
The initial settings were selected from the study
of Huang et al. (2007). In order to optimize the
settings, changes in power (kV and µA), filtration,
degree of rotation and rotation steps, were made,
aiming to optimize the images for the specimen with
ICDAS code 1.
MicroCT imaging of the whole teeth were done
using the microCTSkyscan 1172 from the Indiana
University (IUPUI) (Skyscan 1172, Brussels-
Belgium). To avoid the dehydration of the specimens,
the teeth were wrapped in a thin layer of parafilm
‘M’ (Laboratory Film, American CAN Company,
Greenwich, CT 06830).
Fifteen different scan settings and sixteen
different reconstructions were performed on the three
selected teeth. As the ICDAS 1 code (specimen 2)
was impossible to identify on the first scans, two
different scan settings were performed with ICDAS 2
(specimen 3) aiming the best lesion visualization.
After achieving the best setting for specimen 3, the
same settings were performed with specimen 2,
which allowed identification of the lesion on
specimen 2.
As we were working with high energy
parameters, to achieve the best image quality for
enamel the Al + Cu filter of the Sky scan 1172
equipment was chosen. This specific filter delivered
better diagnostic images and no beam hardening
artifacts were detected with this filtration.
Reconstruction was performed with 100% of beam
hardening correction.
To calibrate the scans, the Scan co Phantom was
used and also a phantom constructed in house. The
Scan co phantom is a marketed unit, made of
hydroxyapatite (HA) of known densities: 0.000 (resin
only), 0.099, 0.199, 0.399, and 0.800 g x cm-3
of HA.
With the higher density areas of the phantoms used,
the corrections were made and the calibration was
performed.
Undersampling was not used to reconstruct the
images because it might cause view aliasing artifacts,
as described before.
It was chosen maximum ring artifact reduction on the
reconstruction software of the Skyscan apparatus.
As identification of early enamel lesions (code 1) is
likely to be the most challenging, after the initial
scans, it was decided to determine the optimal
settings, both for acquisition and for reconstruction
on one of the selected teeth with ICDAS code 1
(Specimen 2). Three different acquisition settings
varying the rotation of the scan and beam power, and
four different reconstructions for each acquisition,
varying the post alignment, beam hardening
correction and threshold of contrast were performed,
aiming to achieve optimal image for lesions ICDAS
code 1.The selected settings were the ones with
greater differences between the attenuation
coefficient on sound enamel and on the lesion as
detected by the Data Viewer software (v. 1.3.2
http://skyscan.be/products/downloads.htm). Visually
the selected settings were considered to be the easiest
to recognize the initial lesion as well. The settings
considered optimal were, resolution of 2K, kV 80,
µA 124, pixel size of 6 µm, rotation step of 0.2,
frame average of 4 and 360 degrees of rotation. The
best reconstruction was ring artifact reduction of 20,
beam hardening of 100%, with a threshold of 0,010
to 0,070. All images were stored in a tiff (16 bit) file.
Tables 1 and 2 present all last tested settings,
performed on specimen 2 (ICDAS code 1):
TABLE 1 – ACQUISITION SETTINGS:
Acquisition Resolution Filter kv µA Rotation Step Frame average Rotation
1.1a 2K Al+Cu 80 124 0.2 4 360
1.1b 2K Al+Cu 80 124 0.2 4 180
1.1 c 2K Al+Cu 100 100 0.2 4 360
TABLE 2 – RECONSTRUCTIONS SETTINGS:
Reconstructions Ring Artifact Reduction Beam Hardening Thresholds Storage
Rec 20 88% 0 x 0.095 16 Tiff
Rec2 20 100% 0 x 0.095 16 Tiff
Rec3 20 100% 0 x 0.110 16 Tiff
Rec4 20 100% 0.010 x 0.070 16 Tiff
III. Discussion:
Some hard tissue studies use micro CT as gold
standard for some measurements27,28
. This is still not
well defined and accepted for dental hard tissues.
Although micro CT is increasing its role on the
analysis of enamel and dentin, it must have an
optimal performance, especially if the goals of

observation are slight differences in mineralization,
as initial caries lesions.
To achieve the best settings for dental hard
tissues, microCT should be performed withbeam
hardening correction. To minimize the effects of
beam hardening, several steps can be taken during the
acquisition and also in reconstruction as: selection of
the appropriate scan field of view; the use of an
appropriate filter; use of beam hardening correction
during reconstruction17
.In our study this was initially
minimized with the Al + Cu filtration. This filter
brought better images regarding the quantity of
energy used. The selection of the appropriate field of
view was acquired with the 2K image, that produced
at the end an image of 6 µm of resolution. Also in the
reconstruction it was applied beam hardening
correction from the reconstruction software indicated
for Skyscan 1172.
Calibration Correction is frequently made with
phantoms in a range of sizes to allow the detectors to
be calibrated with compensation for the beam
hardening effects of different tissue of the
specimen17
. In our study we used the Scanco
phantom, but the enamel is considerably harder than
bone and this phantom is more indicated to bone
specimens calibration. So we also tested natural teeth
before performing microCT best settings.
It is also important to perform flat field
optimization to correct the images recorded by the
detectors, not only with the phantoms but with each
specimen scanned. Flat field optimization helps to
calibrate the detectors avoiding beam hardening
effects17
.
We tried to use higher energies (100kV and
100mA) aiming that small differences in
mineralization could be better visualized, but this did
not happen. Images with 84kV and 124mA were
better to analyze, showing good radiopaque
differences on densities for enamel and dentin.
The characteristic that brought a lot of difference
on the image interpretation was the 3600
of rotation
during acquisition. Definition was much better than
with 1800
of rotation. The acquisition in 1800
should
give an image projected as a mirror, but the best
quality of the 3600
images proved that information is
being missed when the acquisition is limited to 1800
.
Although maximum ring artifact reduction was
chosen during the reconstruction, in some scans this
artifact was still observed.
The threshold chosen was the one that delivered
the highest differences between the lesion visualized
and the sound enamel. It can be noticed that small
lesions analyzed were the ICDAS code 1, that it can
be observed clinically when dried for 5 seconds.
Although it is a lesion that the examiner should be
trained to recognize it, we might remember that it is
also clinically visible. It is still not known if microCT
has the ability to recognize such small lesions, not
clinically visible.
It is also important to note that these settings
may not be fully generalized across all microCT
scanners and specimens29
. Each scanner has its own
physical characteristics and a proper software to
perform its reconstruction.
References
[1] Burghardt AJ, Kazakia GJ, Laib A,
Maumdar S. Quantitative Assessment of
Bone Tissue mineralization with
Polychromatic Micro-Computes
Tomography. Calcif Tissue Int, 83, 2008,
129-138.
[2] Zou W, Gao J, Jones AS, Hunter N, Swain
MV. Characterization of a novel calibration
method for mineral density determination of
dentine by X-ray micro-tomography.
Analyst, 134(1),2009 ,72-79.
[3] Boyd SK, Moser S, Kuhn M, Klinck RJ,
Krauze PL, Müller R, Gasser JA. Evaluation
of three-dimensional image registration
methodologies for in vivo micro-computed
tomography.Ann Biomed Eng, 34(10), 2006,
1587-1599.
[4] Mercer CE, Anderson P, Davis
GR.Sequential 3D X-ray microtomographic
measurement of enamel and dentine ablation
by an Er:YAG laser. Br Dent J, 194(2),
2003, 99-104.
[5] RühliFJ, Kuhn G, Evison R, Müller R,
Schultz M. Diagnostic value of micro-CT in
comparison with histology in the qualitative
assessment of historical human skull bone
pathologies. Am J PhysAnthropol, 133(4),
2007, 1099-1111.
[6] Nazarian A, Snyder BD, Zurakowski D,
Müller R. Quantitative micro-computed
tomography: A non invasive method to
assess equivalent bone mineral density.
Bone, 49, 2008, 302-311.
[7] Cioffi I, VanRuijven LJ, Renders GA,
Farella M, Michelotti A, VanEijden TM.
Regional variations in mineralization and
strain distributions in the cortex of the
human mandibular condyle. Bone, 41(6),
2007, 1051-1058.
[8] Davis GR, Wong FS. X-ray micro
tomography of bones and teeth.
PhysiolMeas, 17(3), 1996, 121-146.
[9] Wong F S L, Elliott J C, Anderson A and
Davis GR. X-ray microtomographic study of
the mineral content and structure of a mouse
mandible.J Dent Res, 70, 1991, 691.
Davis G R. Mineral concentration in rat
femoral diaphyses measured by x-ray

microtomography. Calcif. Tissue Int, 56,
1995, 62-70.
[11] Efeoglu N, Wood D, Efeoglu C.
Microcomputerised tomography evaluation
of 10% carbamide peroxide applied to
enamel. J Dent, 33(7), 2005, 561-567.
[12] Huang TTY, Jones AS, He LH, Darendeliler
MA. Characterisation of enamel white spot
lesions using X-ray micro-tomography.
Journal of Dentistry, 35, 2007, 737-743.
[13] Willmott NS; Wong FSL; Davis GR. An X-
ray Microtomographiy Study on the Mineral
Concentration of Carious Dentine Removed
during Cavity Preparation in Deciduous
Molars. Caries Res, 41, 2007, 129-134.
Davis G R. Three dimensional mineral
distribution in the enamel of a rat incisor
measured by x-ray microtomography. Bone,
16, 1995, 690.1995b;16:690.
[15] Wong FSL, Anderson P, Fan H, Davis GR.
X-ray microtomographic study of mineral
concentration distribution in deciduous
enamel. Archives of Oral Biology, 49, 2004,
979-944.
[16] Schweizer S, Hattendorf B, Schneider P,
Aeschlimann B, GaucklerL, Müller R,
Günther D. Preparation and characterization
of calibration standards for bone density
determination by micro-computed
tomography. Analyst, 132(10), 2007, 1040-
1045.
[17] Barrett JF, Keat N. Artifacts in CT:
Recognition and Avoidance. Radiographics,
24(6), 2004, 1679-1691
[18] KyriakouY; Prell D and Kalender WA. Ring
artifact correction for high-resolution micro
CT. Phys Med Biol, 54, 2009, 385-391.
[19] Zandoná AF, Zero DT. Diagnostic tools for
early caries detection. J Am Dent Assoc,
137, 2006, 1675-1684.
[20] Pretty I. Caries detection and diagnosis:
Novel technologies. J Dent, 34(10), 2006,
727-739.
[21] Soviero VM1, Leal SC, Silva RC, Azevedo
RB. Validity of MicroCT for in vitro
detection of proximal carious lesions in
primary molars. J Dent, 40(1), 2012, 35-40.
[22] de Souza AL1, Leal SC, Chaves SB,
Bronkhorst EM, Frencken JE, Creugers NH.
The Caries Assessment Spectrum and
Treatment (CAST) instrument: construct
validation. Eur J Oral Sci, 122(2), 2014,
149-153.
[23] Ismail AI, Sohn W, Tellez M, Sen A,
Amaya A. The International Caries
Detection and Assessment System (ICDAS):
an integrated system for measuring dental
caries. Community Dent Oral Epidemiol, 34,
2006, 1-9.
[24] Jablonski-Momeni A, Stachniss B., Ricketts
D. N., Heinzel-Gutenbrunner M. and Pieper
K. Reproducibility and Accuracy of the
ICDAS-II for Detection of Occlusal Caries
in vitro. Caries Res, 42, 2008, 79-87.
[25] Ismail AI, Sohn W, Tellez M, Sen A,
Amaya A. The International Caries
Detection and Assessment System (ICDAS):
an integrated system for measuring dental
caries. Community Dent Oral Epidemiol, 34,
2006, 1-9.
[26] Ismail AI, Sohn W, Tellez M, Amaya A,
Sen A, Hasson H, Pitts NB. The
International Caries Detection and
Assessment System (ICDAS): an integrated
system for measuring dental caries.
Community Dent Oral Epidemiol, 35, 2007,
170-178.
[27] Rollano-Hijarrubia E; Stockking R; van der
Meer F; Niessen WJ. Imaging of small high-
density structures in CT: a phantom
study.AcadRadiol, 13, 2006, 893-908.
[28] Postnov AA; Vinogradov AV; VanDyck D;
Saveliev SV; De Clerck NM. Quantitative
analysis of bone mineral content by X-ray
microtomography.PhysilMeas, 24, 2003,
165-178.
[29] Burghardt AJ; Kazakia G; Laib A; Majumar
S. Quantitative Assessment of bone tissue
mineralization with Polychromatic micro-
computed tomography. Calcif Tissue Int, 83,
2008, 129-138.

Micro CT settings for caries detection: how to optimize

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Micro CT settings for caries detection: how to optimize (20)

Recently uploaded (20)

Micro CT settings for caries detection: how to optimize