Volumetric Bone Measurement Around Dental

Implants Using 3D Image Superimposition:

A Methodological and Clinical Pilot Study
Eduardo Aydos Villarinho, DDS, MSc, PhD1/André Correia, DDS, MSc, PhD2/Alvaro Vigo, Math, MSc, PhD3/
Nuno Viriato Ramos, Eng, MSc4/Mário Augusto Pires Vaz, Eng, MSc, PhD5/
Rosemary Sadami Arai Shinkai, DDS, MSc, PhD6

Purpose: This study describes the development of a methodology for using three-dimensional
(3D) image superimposition to measure volumetric changes in bone level around dental implants
in comparison with linear measures. Materials and Methods: The sample was comprised of 46
dental implants of 6-mm length and 4.1-mm diameter placed in the posterior maxilla and posterior
mandible in 20 patients. All implants received screw-retained single crowns. Radiographic images
were taken using cone beam computed tomography (CBCT) and digital periapical radiography
after implantation and after 12 and 24 months of functional loading (after crown placement).
Tridimensional reconstructions of the bone perimeter closest to the implant were developed,
superimposed, and volumetrically measured. Linear measures of bone levels were recorded in
periapical radiography images. A multilevel regression model tested volumetric and linear bone loss.
Results: The mean peri-implant linear bone loss for the first and second years was 0.2 ± 0.4 mm
and 0.1 ± 0.2 mm, respectively, and the mean volumetric bone loss for the first and second years
was 7.2 ± 6.1 mm3 and 6.4 ± 7.8 mm3, respectively. It was estimated that an increase of 1 mm
of linear bone loss was associated with a mean volumetric bone loss of approximately 14 mm3
(P < .001). Conclusion: The findings showed that linear and volumetric bone loss measures are
related. Measuring volumetric bone changes around implants is possible provided that the CBCT
images have proper contrast and sharpness, particularly around the implant outline. Improvements
in image quality and in the filters for bone tissue detection would be important for this methodology
to be made faster and used clinically. Int J Prosthodont 2018;31:23–30. doi: 10.11607/ijp.5366

T he long-term success of implant treatment de-

pends on the sustained integrity of the bony in-
terfacial response to dental implants.1,2 Originally
proposed success criteria2 (no clinically observable
movement, no peri-implant radiolucency, vertical bone
loss < 2 mm in the first year and < 0.2 mm in sub-
sequent years, and no persistent signs or symptoms,
1Postdoctoral Fellow, Postgraduate Program in Dentistry, Pontifical Catholic
University of Rio Grande do Sul (PUCRS), Porto Alegre, Brazil.
2Professor, Dental Medicine Degree, Institute of Health Sciences,
Portuguese Catholic University (UCP), Viseu, Portugal.
3Professor, Department of Statistics (Institute of Mathematics and
Statistics), Federal University of Rio Grande do Sul, Porto Alegre, Brazil.
4Researcher, Science and Innovation Institute of Mechanical Engineering
(INEGI), Porto, Portugal.
5Professor, Department of Mechanical Engineering, Faculty of Engeneering,
University of Porto, Portugal.
6Professor, Postgraduate Program in Dentistry, Pontifical Catholic University
of Rio Grande do Sul (PUCRS), Dental School, Porto Alegre, Brazil.
Correspondence to: Dr Eduardo Aydos Villarinho, Avenida Ipiranga,
6681 – Predio 6. CEP: 90619-900. Porto Alegre, RS, Brazil.
Email: eduardo.villarinho@acad.pucrs.br
©2018 by Quintessence Publishing Co Inc.
such as pain, infection, neuropathy, paresthesia, and
injury to the mandibular canal) are still commonly em-
ployed, with periapical radiographs commonly used
to monitor changes in bone levels around dental im-
plants. However, these radiographic images contain
distortions due to intrinsic geometric inaccuracies of
the exam and the difficulty of using the paralleling
technique according to the patient’s anatomical con-
ditions. In addition, two-dimensional (2D) periapical
radiographs generate image superimposition, making
it impossible to differentiate the buccal bone from the
palatal or lingual bone; thus, periapical radiographs
provide a linear analysis of bone loss of the mesial and
distal surfaces and cannot be used to assess the buc-
cal, lingual, or palatal surfaces. Volumetric measure-
ment is also impossible, as these measurements can
only be done using tests that allow three-dimensional
(3D) volumetric image reconstruction, such as in com-
puted tomography (CT) scanning.3–5
The increasing use of cone beam computed tomog-
raphy (CBCT) in dentistry has resulted in significant
improvements in diagnosis and treatment plan de-
velopment in different fields. CBCT provides imaging
with volumetric reconstruction with better contrast

Volume 31, Number 1, 2018 23

© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Volumetric Bone Measurement Around Dental Implants Using 3D Image Superimposition
Table 1   D
 escription of the Demographic and Clinical
Sample Data
Variable
Patients (n) 20
 Female 8 sample of a prospective cohort study with clinical
 Male 12
 Age (y) (mean ± SD) 52 ± 10
Implants (n) 46
  In maxilla 23 (16 in molar and 7 in premolar region)
  In mandible 23 (18 in molar and 5 in premolar region)
and spatial resolution than multi-slice computed to-
mography (MSCT) and allows for distinguishing ob-
jects separated by subtle differences in distance and
radiopacity, which is ideal for detecting and visualiz-
ing peri-implant defects.6–8
Previous clinical implant follow-up studies with CT
have assessed bone level changes; however, only lin-
ear changes in buccal and palatal bone plates were
analyzed—which would not be possible with periapi-
cal radiography—and do not utilize the volumetric
reconstruction capability of CT to analyze volumetric
changes in the bone tissue adjacent to the implant
during a given period of time.9–12 In vitro and animal
studies on volumetric measurement of bone defects
using CT have shown its potential to provide such
measurements.3,7,8,13
Superimposition of different 3D images obtained by
CBCT is a method that allows for analyzing the longi-
tudinal evolution of a given treatment, providing volu-
metric changes in a given area of interest. In dentistry,
this method has been used for monitoring mandibular
advancement,14,15 condylar remodeling in orthogna-
thic surgery patients,16–19 and synthetic grafting in
the malar region, along with orthognathic surgery20
in segmented alveolar areas that have been subject-
ed to different alveolar preservation methods with or
without implant placement.21 In 2013, Ahmad et al22
monitored the bone remodeling process through 3D
image superimposition in patients rehabilitated with
implant-retained mandibular overdentures in the ca-
nine (cuspid) region, comparing the images from be-
fore the surgery and at 12 and 24 months after the
prosthesis had been in function. However, analysis of
dimensional changes was carried out in the most pos-
terior region—where the prosthesis was anchored—
and not in the region close to the implants, which may
be explained by the presence of artifacts in the im-
plant region.
Therefore, this pilot study aimed to describe the
development of a new methodology for measuring
volumetric changes in bone levels around osseoin-
tegrated implants using 3D image superimposition
and compared this new method with conventional
radiography.
24 The International Journal of Prosthodontics
© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Methodology
This work was conducted as a methodological de-
velopment pilot study. Data were obtained from the
and radiographic follow-up,23 which was approved
by the Institutional Review Board and registered
at the National System of Information on Ethics for
Research in Humans. The study sample consisted of
46 Standard Plus Regular Neck SLActive implants
(Straumann Dental Implant System) of 6-mm length
and 4.1-mm diameter placed in a nonprobabilistic
sample of patients (Table 1) following the one-stage
surgical protocol, as recommended by the manufac-
turer. After a 90-day osseointegration period, transfer
molding was carried out for manufacture of the met-
al-ceramic crown. SynOcta abutments and castable
cylinders (Straumann Dental Implant System) were
used to fabricate screw-retained crowns.
The exclusion criteria were: Previous episodes of
failure of osseointegration at the region of interest,
uncontrolled type 2 diabetes, immunosuppression,
head and neck radiotherapy, active periodontal dis-
ease at the remaining dentition, poor oral hygiene,
and use of a removable partial denture (RPD) or com-
plete denture (CD) in the opposing dentition.
Radiographic images were taken using CBCT and
digital periapical radiography. CBCT images were
acquired in Digital Imaging and Communications in
Medicine (DICOM) format by CBCT scanner i-CAT
CBCT (Imaging Sciences International) and KODAK
9500 Cone Beam 3D System (Kodak Dental Systems,
Carestream Health) at two different clinics specialized
in dental imaging diagnostics. Baseline periapical ra-
diographic images were obtained immediately after
surgery, and baseline CBCT images were acquired
within 1 week after surgery. Follow-up radiographic
and CBCT images were taken at 12 and 24 months
after the prosthesis was in function.
Digital periapical radiographs were taken using the
long cone paralleling technique, with x-ray holders
(Rinn XCP) customized with silicone putty for each
patient. The Krystal X Easy (Owandy) digital sensor
and its software (Owandy Quick Vision) were used
for image acquisition. The Timex 70C (Gnatus, Brazil)
x-ray equipment was used, operating at 65 kVp and
7.5 mA and using exposure times of 0.10 seconds (for
maxillary and mandibular premolars and mandibular
molars) and 0.12 seconds (maxillary molars).
Image Treatment and Measurement Procedures
Image 3D reconstruction was performed using
Mimics (Materialise), a software for medical image
processing, and all DICOM images were imported

a b
c d
Fig 1   Mimics image with bone region selected around the implant,
checking that only the bone tissue is selected in all sections, and sub-
sequent 3D reconstruction. (a) Coronal section. (b) Axial section.
(c) Sagittal section. (d) 3D reconstruction of the selected area.
a b
Fig 3  (a) Bone perimeter around selected implant after removal of
area of no interest for measurement. (b) Image during mesh refine-
ment for better image superimposition.
and processed. After uploading different sections, the
thresholding tool was applied. In bone scale mode,
the area of interest around the implant was selected,
and then the image-processing tool crop mask was
applied. Within the selected area, each section was
individually examined in the sagittal, axial, and coro-
nal planes, ensuring that only the bone tissue around
the implant being assessed was highlighted in each
section. The processing of images of a single implant
at different times was performed in the same man-
ner by one trained operator, ensuring that all recon-
structions followed the same standards. After image
manipulation was completed, a 3D model of selected
peri-implant bone tissue was generated and saved in
Standard Tessellation Language (STL) format (Fig 1).
The STL model was exported to Geomagic Studio
(Geomagic), and the area of interest (ie, the bone pe-
rimeter closest to the implant) was segmented (Fig 2).
Then, the image mesh was refined, peaks were re-
moved, polygons were relaxed, and minor imperfec-
tions were manually corrected (ie, filled up) (Fig 3).
Different image-processing routines were repeated
for all time points for a given single implant. Then, im-
ages were superimposed using an alignment tool with
semi-automatic registration wizard. Using a color map,
© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Villarinho et al
a b
Fig 2   Stereolithography (STL) image of selected bone tissue import-
ed into Geometics. (a) Original imported image. (b) Selected image
with closest perimeter to selected implant.
Fig 4   Images at different times, superimposed by alignment tool. The
blending of colors in the bottom image shows that superimposition is
being correctly made.
Fig 5  Correct im-
age alignment is veri-
fied through a color
map, where green
color means that im-
ages for the bone
perimeter closest to
the implant are per-
fectly aligned or at a
distance < 0.041 mm.
images were visually inspected to check the correct
superimposition, showing areas of greater contact and
greater distance between images (Figs 4 and 5).
With this set of superimposed images of a single im-
plant, external walls were projected onto the selected
area, creating a cube around the point of implant in-
sertion. In this cube, external walls were the same for
all images. As the cube walls were the same for all im-
ages at different times, difference in cube volume was
due to the bone portion closest to the implant, which
Volume 31, Number 1, 2018 25
Volumetric Bone Measurement Around Dental Implants Using 3D Image Superimposition

The linear bone loss was measured using ImageJ

software, version 1.3V (free software from Wayne
Rasband, National Institutes of Health). The mea-
surements were performed for the mesial and distal
surfaces, and the means of the two surfaces were
calculated.23
Two measurements were recorded per implant by
the same blinded examiner, with a minimum interval
of 15 days (intra-examiner agreement was assessed
using intraclass correlation coefficient; ICC = 0.9).

Statistical Analysis

Descriptive analyses of the data on volumetric and

linear changes in the peri-implant tissue were per-
Fig 6   Images with cube walls already projected, where external di- formed to characterize the sample. To analyze bone
mensions are the same for all times, and illustrating a possible cube loss over time, a multilevel model with repeated mea-
section.
sures was used to evaluate the effect of the predictor
(linear bone loss) on volumetric bone loss consider-
ing the hierarchical implant structure in each patient
and correlation structure (exchangeable). Data were
Table 2   D
 escriptive Results of Volumetric and Linear analyzed using the PROC NLMIXED procedure in
Bone Loss by Period SAS Version 9.4.24
0–1 y 1–2 y 0–2 y
Linear bone loss (mm) Results
 Mean maxilla and 0.2 ± 0.4 0.1 ± 0.2 0.2 ± 0.3
mandible (–0.4; 1.6) (–0.7; 0.4) (–0.6; 0.8) The original sample was comprised of 45 implants. Two
n = 45 n = 38 n = 38
were lost between the first and second years (sub-
 Maxilla 0.1 ± 0.3 0.1 ± 0.2 0.2 ± 0.3
(–0.3; 0.5) (–0.7; 0.4) (–0.2; 0.7) tracting two implants in the analysis between years
n = 23 n = 20 n = 20 1 and 2), and eight implants showed only one tomo-
 Mandible 0.3 ± 0.5 0.1 ± 0.1 0.1 ± 0.4 graphic image at one of the follow-up times, making it
(–0.4; 1.6) (–0.2; 0.4) (–0.6; 0.8) impossible to make comparative measurements (sub-
n = 22 n = 18 n = 18
tracting eight implants from the overall analysis). Four
Volumetric bone loss (mm³)
tests of different implants (containing six implants at
 Mean maxilla and 7.2 ± 6.1 6.4 ± 7.8 12.6 ± 8.0
mandible (–0.5; 23.8) (–7.3; 29.0) (–1.7; 20.8) baseline and 1 at the first year) were disregarded due
n = 30 n = 34 n = 28 to low image sharpness, which did not allow superim-
 Maxilla 5.2 ± 5.1 7.3 ± 6.0 13.9 ± 6.0 position with the other tests at different times (ie, sub-
(–0.5; 15.0) (–4.1; 16.3) (1.7; 26.3) tracting seven implants between baseline and year 1,
n = 13 n = 16 n = 13
and one implant between years 1 and 2). One implant
 Mandible 8.7 ± 6.5 5.6 ± 9.1 11.5 ± 9.3
(1.1; 23.8) (–7.3; 29.0) (–1.7; 20.8) was not considered for final analysis because, as its
n = 17 n = 18 n = 15 surrounding tissues were surgically modified during
Data are reported as mean ± standard deviation (SD) and range follow-up due to extraction of an adjacent tooth and
(minimum; maximum). subsequent grafting and implant placement, it was an
outlier and significantly altered averages and devia-
tions. Therefore, volumetric changes were measured
varied from one image to another. Bone volumetric in 30 implants during the first year of function and in
variation was calculated in mm3 around implants at 34 implants during the second year of function, and
different periods. Cubes were saved through axis ori- accumulated changes during both years were mea-
entation, similar to that imported from Mimics, so the sured in 28 implants.
mesial, distal, buccal, and lingual or palatal surfaces Volumetric and linear changes in peri-implant bone
could be identified (Fig 6). After surface identifica- levels at different times are described in Table 2.
tion, cubes were available to be segmented in sev- The multilevel regression model estimated that an
eral ways, which made it possible to examine different increase of 1 mm of linear bone loss was associated
faces of peri-implant tissue in isolation according to with a mean volumetric bone loss of approximately
the examiner’s interest. 14.02 mm3 (P < .001) (Table 3).

26 The International Journal of Prosthodontics

© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.

Table 3   M
 ultilevel Regression Model to Evaluate Volumetric Changes in Bone Level Around Dental Implants in
Relation with Linear Measures
Volumetric bone loss (mm3)
Parameter Estimate Standard error
Intercept 183.5758 8.7877
Linear bone loss 14.0200 3.2071
Predictor: Linear bone loss (1 mm).
Discussion
This study describes a new method for using CBCT
image superimposition to determine volumetric bone
loss around dental implants. The results showed an
average volumetric bone loss of 7.2 ± 6.1 mm3 for the
first year of function and 12.6 ± 8.0 mm3 during the
accumulated period of 2 years of function and esti-
mated that an increase of 1 mm of linear bone loss
was associated with a mean volumetric bone loss of
approximately 14.02 mm3. This data suggests linear
behavior and tissue stability over time. Bone loss was
similar in the first and second years of implant func-
tion, as also shown in previous studies.1,2 This fact
could be explained by the implant system used, which
has the implant-prosthesis interface vertically away
from the bone tissue, where initial bone remodeling
does not occur after prosthesis placement or after the
second surgical stage.
Results on volumetric bone changes around im-
plants at different follow-up times cannot be directly
compared to existing data in the literature, as this is
a new 3D measurement method. Even in clinical and
tomographic implant follow-up studies, 9,10,12 longi-
tudinal images are used for linear measurements at
different time points. When image superimposition
is used, it is a 2D image superimposition10,12 of sec-
tions selected in the CBCT rather than in volumetric
reconstructions.
This new approach using superimposition of STL
images obtained by CBCT can be an objective and
reproducible method for 3D quantification of bone
remodeling.22 However, in the case of bone regions
around dental implants that have metal objects and
teeth, superimposition becomes more critical. The
anticipated volumetric change tends to be very small
during the period, requiring extremely accurate im-
age superimposition and tests with good contrast and
sharpness, which is not always possible. Conversely,
superimposition using the implant format allows this
technique to be applied in any clinical situation, as it
does not rely on other anatomical references, such as
adjacent teeth, to align images. In addition, once im-
ages are superimposed, there are a wide variety of
options for 3D analysis. Images of different times are
segmented at once, ensuring that the same area of
© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Villarinho et al
95% Confidence limits P
166.3523 200.7994 < .0001
20.3059 7.7341 < .0001
interest is analyzed in all tests. In the present study,
this allowed the bone region of interest to be divided
into mesial and distal portions, as well as buccal and
palatal or lingual portions, for analysis.
For this 3D image superimposition methodology,
good image quality is crucial—particularly sharpness
of implant edges—as it is through outlines that differ-
ent images are superimposed. Ahmad et al22 did not
evaluate tissue changes around implants supporting
mandibular overdentures due to the presence of arti-
facts in the superimposed CBCT images and assessed
only the posterior ridge region, where the prosthesis
was anchored. Unfortunately, unlike periapical radio-
graphs, the presence of metal objects in CBCT—such
as amalgam restorations, cast inlays, metal crowns,
and implants—can generate artifacts, reducing image
quality. Artifacts are shown as hyperdense white lines
radiating from the metal object or dark areas around
the metal object, where no image information is avail-
able.3,6,7,13,26 Interference of these artifacts on image
quality was reported to be higher for CBCT than for
MSCT by Draenert et al.27
In this study, artifacts appearing as hyperdense
lines and spots did not allow definition of the implant
edges or image superimposition, making it impossible
to perform volumetric measurement. On the other
hand, artifacts appearing as dark spots did not pre-
vent measurement but made it difficult to manipulate
images. Kamburoglu et al8 suggested that artifacts
could limit the use of CBCT for identification of de-
fects in the buccal plate; however, the authors showed
good defect detection and measurement efficiency.
For bone tissue selection, the image-processing
tool threshold value was applied individually for each
test, with section-by-section manual verification of
adding or removing selected areas. Ahmad et al22 also
used the threshold value filter individually for each
test, although they did not report changes in sections
when done manually. They conducted a pilot study
and concluded that changes in the threshold value
of up to 20% do not significantly affect STL recon-
structions. In this study, however, images presented
artifacts possibly from dental implants, metal-ceramic
crowns, metal restorations, and even teeth, which
could be confused with bone tissue without the use
of specific filters.
Volume 31, Number 1, 2018 27
Volumetric Bone Measurement Around Dental Implants Using 3D Image Superimposition
Conventional and radiographic images are still com-
plementary in clinics. Vandenberghe et al28 compared
digital periapical radiographic images with CBCT im-
ages using a corpse’s jaw and a dry skull as the gold
standard and concluded that both methods are ef-
ficient to determine the defect depth, presenting the
same rates for defect underestimation and overesti-
mation. However, they observed that the lamina dura,
trabecular bone characteristics, and contrast among
tissues are best visualized in periapical radiographs,
while defects in the furcation area and the descrip-
tion of defect morphology are best visualized by to-
mography. Thus, they concluded that one test does not
replace the other in every aspect and that, in some
cases, both tests are required for accurate testing.
In another comparative study, Corpas et al3 analyzed
bone levels around implants to compare periapical ra-
diographs and CBCT in an animal model, having histo-
logic sections as a gold standard. For both techniques,
the image and histologic measurements were strongly
correlated; however, both tests underestimated actual
values (variation was 1.17 mm for periapical radio-
graphs and 1.20 mm for CT scans). The researchers
point out, however, that 50% of the deviations were
lower than 0.5 mm, which is clinically insignificant.
The literature shows that CBCT is efficient for di-
agnosis and measurement of peri-implant bone de-
fects.3,8,13,27 Additionally, only tomographic scans,
which allow for 3D reconstruction, enable visualization
of such defects in the buccal, palatal, or lingual bone
plate, as well as follow-up of changes in bone levels in
these regions over time. Mengel et al13 proved the ac-
curacy of CBCT for measuring bone defects, reporting
that CBCT was even more accurate than MSCT.
CBCT provides clinical information that would not
be possible using conventional radiography. However,
its use should be judicious because of the potential
risks related to ionizing radiation. In a literature re-
view, Lorenzoni et al29 reported that MSCT showed
the highest levels of radiation, followed by CBCT
and conventional radiography. They reported that
increased kV, mA, exposure time, and field of view
(FOV) led to increased radiation doses, and a numeric
dose comparison between the different scans would
not be possible because the number varies depend-
ing on the equipment used and its adjustment during
image acquisition. Li30 reported that CBCT radiation
doses could be on average 10 times lower than con-
ventional tomography, but also 100 times higher than
conventional radiography.
To compare the mutagenicity and cytotoxicity of
ionizing radiation in CBCT and conventional radiog-
raphy, Lorenzoni et al31 reported increased nuclear
changes in patients after both tests. CBCT presented
a higher number of changes and cell death; however,
28 The International Journal of Prosthodontics
© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
the increase in micronuclei was not significant, and
mutagenicity (which would cause permanent dam-
age to patients subjected to the tests) did not occur.
These results corroborate the findings of Silva et al,32
who analyzed cell changes in patients subjected to
one or two panoramic radiographs and showed that
increased radiation led to increased cell changes but
mutagenic changes did not occur (ie, there was no
permanent damage). In any case, when selecting the
type of scanning, as well as the equipment and its
adjustment, the principle of minimum-required dose
should be observed for correct and accurate diagnos-
tic imaging.
This pilot study presents only the use of 6-mm
implants in the posterior regions of the maxilla and
mandible, where, despite having a greater magnitude
of masticatory forces, the incidence of forces is pre-
dominantly axial. It is worth noting the importance of
studies with a greater number and variety of implants
located in different regions (eg, in the anterior re-
gions, where a greater incidence of oblique forces is
expected and a different tissue response is therefore
potentially generated).
In summary, this investigation developed a prom-
ising methodology for volumetric bone measurement
around dental implants. However, as medical image-
processing programs cannot properly distinguish tis-
sues automatically, this methodology still depends
on the examiner, making the process relatively slow,
subjective, and dependent on adequate training. This
methodology requires high tomographic image qual-
ity to ensure that small tissue changes are not from
artifact-related errors and are instead from anatomi-
cal changes in the area being studied. Improvements
in image quality and in the filters for bone tissue de-
tection would be important for this methodology to
become faster and able to be used clinically. These
issues should be addressed in further studies to im-
prove the technical procedures and to allow the use
of this methodology for several clinical situations that
might benefit from 3D measurement of bone level
changes over time.
Conclusions
Based on the findings and limitations of this study, it
can be concluded that:
•• Measuring volumetric bone changes around
implants is possible provided that CBCT images
have good contrast and sharpness, particularly in
the implant outline.
•• It was estimated that an increase of 1 mm of linear
bone loss was associated with a mean volumetric
bone loss of approximately 14.02 mm3.

Acknowledgments
The authors would like to thank Dr Diego Triches, Dr Fernando
Alonso, Dr Luis Mezzomo, and Dr Eduardo Rolim Teixeira, team
members of the clinical research; Dr Rodrigo Miller and Dr
Mauricio Peixoto for their help during clinical data collection; and
Dr Gustavo Barbosa, for his administrative research assistance.
The clinical project received financial support from the Brazilian
Ministry of Education/Coordination for the Improvement of Higher
Education Personnel (CAPES), the Brazilian Ministry of Science,
Technology and Innovation/National Counsel of Technological
and Scientific Development (CNPq), and the International Team
for Implantology (ITI). The authors declare that there are no con-
flicts of interest in the present study.
References
  1. Adell R, Lekholm U, Rockler B, Brånemark PI. A 15-year study
of osseointegrated implants in the treatment of the edentulous
jaw. Int J Oral Surg 1981;10:387–416.
  2. Albrektsson T, Zarb G, Worthington P, Eriksson AR. The long-
term efficacy of currently used dental implants: A review and
proposed criteria of success. Int J Oral Maxillofac Implants
1986;1:11–25.
  3. Corpas Ldos S, Jacobs R, Quirynen M, Huang Y, Naert I, Duyck
J. Peri-implant bone tissue assessment by comparing the
outcome of intra-oral radiograph and cone beam computed
tomography analyses to the histological standard. Clin Oral
Implants Res 2011;22:492–499.
  4. Hermann JS, Schoolfield JD, Nummikoski PV, Buser D, Schenk
RK, Cochran DL. Crestal bone changes around titanium im-
plants: A methodologic study comparing linear radiographic
with histometric measurements. Int J Oral Maxillofac Implants
2001;16:475–485.
  5. Tyndall DA, Brooks SL. Selection criteria for dental implant site
imaging: A position paper of the American Academy of Oral
and Maxillofacial Radiology. Oral Surg Oral Med Oral Pathol
Oral Radiol Endod 2000;89:630–637.
 6. Angelopoulos C, Scarfe WC, Farman AG. A comparison of
maxillofacial CBCT and medical CT. Atlas Oral Maxillofac Surg
Clin North Am 2012;20:1–17.
  7. Sirin Y, Horasan S, Yaman D, et al. Detection of crestal radiolu-
cencies around dental implants: An in vitro experimental study.
J Oral Maxillofac Surg 2012;70:1540–1550.
  8. Kamburog˘lu K, Murat S, Kılıç C, et al. Accuracy of CBCT im-
ages in the assessment of buccal marginal alveolar peri-im-
plant defects: Effect of field of view. Dentomaxillofac Radiol
2014;43:20130332.
  9. Miyamoto Y, Obama T. Dental cone beam computed tomog-
raphy analyses of postoperative labial bone thickness in
maxillary anterior implants: Comparing immediate and de-
layed implant placement. Int J  Periodontics Restorative  Dent
2011;31:215–225.
10. Roe P, Kan JY, Rungcharassaeng K, Caruso JM, Zimmerman
G, Mesquida J. Horizontal and vertical dimensional changes
of peri-implant facial bone following immediate placement
and provisionalization of maxillary anterior single implants:
A 1-year cone beam computed tomography study. Int J Oral
Maxillofac Implants 2012;27:393–400.
© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.
Villarinho et al
11. Kaminaka A, Nakano T, Ono S, Kato T, Yatani H. Cone-beam
computed tomography evaluation of horizontal and vertical di-
mensional changes in buccal peri-implant alveolar bone and
soft tissue: A 1-year prospective clinical study. Clin Implant
Dent Relat Res 2015;17(suppl):e576–e585.
12. Morimoto T, Tsukiyama Y, Morimoto K, Koyano K. Facial bone
alterations on maxillary anterior single implants for immediate
placement and provisionalization following tooth extraction:
A superimposed cone beam computed tomography study.
Clin Oral Implants Res 2015;26:1383–1389.
13. Mengel R, Kruse B, Flores-de-Jacoby L. Digital volume tomog-
raphy in the diagnosis of peri-implant defects: An in vitro study
on native pig mandibles. J Periodontol 2006;77:1234–1241.
14. Almeida RC, Cevidanes LH, Carvalho FA, et al. Soft tissue re-
sponse to mandibular advancement using 3D CBCT scanning.
Int J Oral Maxillofac Surg 2011;40:353–359.
15. Maal TJ, de Koning MJ, Plooij JM, et al. One year postoperative
hard and soft tissue volumetric changes after a BSSO mandib-
ular advancement. Int J Oral Maxillofac Surg 2012;41:1137–1145.
16. Kim YI, Park SB, Son WS, Hwang DS. Midfacial soft-tissue
changes after advancement of maxilla with Le Fort I osteotomy
and mandibular setback surgery: Comparison of conventional
and high Le Fort I osteotomies by superimposition of cone-
beam computed tomography volumes. J Oral Maxillofac Surg
2011;69:e225–e233.
17. Kim YI, Park SB, Jung YH, Hwang DS, Lee JY. Evaluation of
intersegmental displacement according to osteosynthesis
method for mandibular setback sagittal split ramus osteotomy
using cone-beam computed tomographic superimposition.
J Oral Maxillofac Surg 2012;70:2893–2898.
18. Park SB, Yang YM, Kim YI, Cho BH, Jung YH, Hwang DS. Effect
of bimaxillary surgery on adaptive condylar head remodel-
ing: Metric analysis and image interpretation using cone-
beam computed tomography volume superimposition. J Oral
Maxillofac Surg 2012;70:1951–1959.
19. Park SB, Yoon JK, Kim YI, Hwang DS, Cho BH, Son WS. The
evaluation of the nasal morphologic changes after bimaxillary
surgery in skeletal class III malocclusion by using the super-
imposition of cone-beam computed tomography (CBCT) vol-
umes. J Craniomaxillofac Surg 2012;40:e87–e92.
20. Grybauskas S, Locs J, Salma I, Salms G, Berzina-Cimdina L.
Volumetric analysis of implanted biphasic calcium phosphate/
collagen composite by three-dimensional cone beam computed
tomography head model superimposition. J Craniomaxillofac
Surg 2015;43:167–174.
21. Economopoulos TL, Asvestas PA, Matsopoulos GK, Molnár
B, Windisch P. Volumetric difference evaluation of registered
three-dimensional pre-operative and post-operative CT dental
data. Dentomaxillofac Radiol 2012;41:328–339.
22. Ahmad R, Abu-Hassan MI, Li Q, Swain MV. Three dimensional
quantification of mandibular bone remodeling using standard
tessellation language registration based superimposition. Clin
Oral Implants Res 2013;24:1273–1279.
23. Villarinho EA, Triches DF, Alonso FR, Mezzomo LAM, Teixeira
ER, Shinkai RSA. Risk factors for single crowns supported by
short (6-mm) implants in the posterior region: A prospective
clinical and radiographic study. Clin Implant Dent Relat Res
2017;19:671–680.
24. Lai HC, Si MS, Zhuang LF, Shen H, Liu YL, Wismeijer D. Long-
term outcomes of short dental implants supporting single
crowns in posterior region: A clinical retrospective study of
5–10 years. Clin Oral Implants Res 2013;24:230–237.
Volume 31, Number 1, 2018 29
Volumetric Bone Measurement Around Dental Implants Using 3D Image Superimposition

25. Chuang SK, Cai T, Douglass CW, Wei LJ, Dodson TB. Frailty
approach for the analysis of clustered failure time observations
in dental research. J Dent Res 2005;84:54–58.
26. Kamburoglu K, Kolsuz E, Murat S, Eren H, Yüksel S, Paksoy
CS. Assessment of buccal marginal alveolar peri-implant and
periodontal defects using a cone beam CT system with and
without the application of metal artefact reduction mode.
Dentomaxillofac Radiol 2013;42:20130176.
27. Draenert FG, Coppenrath E, Herzog P, Müller S, Mueller-Lisse
UG. Beam hardening artefacts occur in dental implant scans
with the NewTom cone beam CT but not with the dental 4-row
multidetector CT. Dentomaxillofac Radiol 2007;36:198–203.
28. Vandenberghe B, Jacobs R, Yang J. Diagnostic validity (or
acuity) of 2D CCD versus 3D CBCT-images for assessing
periodontal breakdown. Oral Surg Oral Med Oral Pathol Oral
Radiol Endod 2007;104:395–401.
29. Lorenzoni DC, Bolognese AM, Garib DG, Guedes FR, Sant’anna
EF. Cone-beam computed tomography and radiographs in
dentistry: Aspects related to radiation dose. Int J Dent 2012;
2012:813768.
30. Li G. Patient radiation dose and protection from cone-beam
computed tomography. Imaging Sci Dent 2013;43:63–69.
31. Lorenzoni DC, Fracalossi AC, Carlin V, Ribeiro DA, Sant’anna
EF. Mutagenicity and cytotoxicity in patients submitted to ion-
izing radiation. Angle Orthod 2013;83:104–109.
32. da Silva AE, Rados PV, da Silva Lauxen I, Gedoz L, Villarinho EA,
Fontanella V. Nuclear changes in tongue epithelial cells follow-
ing panoramic radiography. Mutat Res 2007;632:121–125.

Literature Abstract

Body-Mass Index and Risk of 22 Specific Cancers: A Population-Based Cohort Study of 5·24 Million UK Adults

High body mass index (BMI) predisposes individuals to several site-specific cancers, but a large-scale systematic and detailed
characterization of patterns of risk across all common cancers, adjusted for potential confounders, has not been previously undertaken.
The authors therefore aimed to investigate the links between BMI and the most common site-specific cancers. With primary care and BMI
data from individuals in the Clinical Practice Research Datalink, Cox models were fitted to investigate associations between BMI and 22
of the most common cancers, adjusting for potential confounders. Linear and then non-linear (spline) models were fitted to the data, and
effects of sex, menopausal status, smoking, age, and population were investigated. A total of 5.24 million individuals were included, and
166,955 developed cancers of interest. BMI was associated with 17 of 22 cancers, but effects varied substantially by site. Using a BMI scale
of < 18.5 to > 25, each increase in BMI of 5 kg/m2 was roughly linearly associated with cancers of the uterus (hazard ratio [HR] 1.62; 99%
CI 1.56–1.69; P < .0001), gallbladder (HR 1.31; 99% CI 1.12–1.52; P < .0001), kidney (HR 1.25; 99% CI 1.17–1.33; P <. 0001), cervix (HR 1.10;
99% CI 1.03–1.17; P = . 00035), thyroid (HR 1.09; 99% CI 1.00–1.19; P = .0088), and leukemia (HR 1.09, 99% CI 1.05–1.13; P ≤ .0001). BMI
was positively associated with liver (HR 1.19; 99% CI 1.12–1.27), colon (HR 1.10; 99% CI 1.07–1.13), ovarian (HR 1.09; 99% CI 1.04–1.14),
and postmenopausal breast cancers (HR 1.05; 99% CI 1.03–1.07) overall (all P < .0001), but these effects varied by BMI or individual-level
characteristics. Prostate and premenopausal breast cancer risk showed inverse associations with BMI, both overall (prostate HR 0.98,
99% CI 0.95–1.00; premenopausal breast cancer HR 0.89, 99% CI 0.86–0.92) and in never-smokers (prostate HR 0.96, 99% CI 0.93–0.99;
premenopausal breast cancer HR 0.89, 99% CI 0.85–0.94). By contrast, for lung and oral cavity cancer, no association was observed in
never-smokers (lung HR 0.99, 99% CI 0.93–1.05; oral cavity HR 1.07, 99% CI 0.91–1.26). Overall inverse associations were driven by current
smokers and ex-smokers, probably because of residual confounding by smoking amount. Assuming causality, 41% of uterine and 10%
or more of gallbladder, kidney, liver, and colon cancers could be attributable to excess weight. It was estimated that a population-wide
increase in BMI of 1 kg/m2 would result in 3,790 additional annual UK patients developing 1 of the 10 cancers positively associated with
BMI. BMI is associated with cancer risk, with substantial population-level effects. The heterogeneity in the effects suggests that different
mechanisms are associated with different cancer sites and different patient subgroups.

Bhaskaran K, Douglas I, Forbes H, dos-Santos-Silva I, Leon DA, Smeeth L. Lancet 2014;384:755–765. References: 39. Reprints: Krishnan
Bhaskaran, Krishnan.bhaskaran@lshtm.ac.uk —Steven Sadowsky, USA

30 The International Journal of Prosthodontics

© 2018 BY QUINTESSENCE PUBLISHING CO, INC. PRINTING OF THIS DOCUMENT IS RESTRICTED TO PERSONAL USE ONLY.
NO PART MAY BE REPRODUCED OR TRANSMITTED IN ANY FORM WITHOUT WRITTEN PERMISSION FROM THE PUBLISHER.