İki ve üç boyutlu sefalometrik ölçümler arasındaki farklılıkların değerlendirilmesi

PDF Dosyası: 

arastirmax_1427_pp_43-49.pdf

Introduction
Cephalometric radiographs have been used for diagnosis, treatment planning, and evaluation of treatment results in orthodontics (1-6). Not only the size
and location of the maxilla and the mandible, but the
relationship between the craniofacial structures can
also be defined by the measurements made on these
radiographs. Cephalometric analyses also show the
soft tissue profile, and the positions and relations of
the upper and lower incisors (3,4). Despite the advantages of low cost, low radiation dose, and high reproducibility, lateral cephalograms have several drawbacks especially in the evaluation of facial asymmetry
and in the planning of orthognathic surgery (5-9).
In a conventional orthodontic treatment planning,
the cephalometric analyses are used to measure values of skeletal, dental and soft tissue landmarks,
and obtained values are compared with standard
mean values to diagnose the problems of the patients
(1,4,6). However, ideal skeletal, dental, and soft tissue
relationship can differ according to the cephalometric analysis method used. There is a wide database
regarding radiographic cephalometry since this technique has been used for long years in orthodontics.
Values of clinically normal patients are calculated
and aberrations are determined angularly, linearly,
or proportionally by using these databases so, cephalometric evaluation still maintains its indispensable
way in the orthodontic treatment planning (6).
Recently, new software programs have been developed to enable the analysis of data which are obtained
by three dimensional (3D) visualisation techniques
(1,2,4). Evaluation and measurement of craniofacial
structures by 3D cephalometric analyses, developing
orthodontic treatment planning, post-treatment soft
tissue simulations, and real 3D solid biomodelling
have been possible by these techniques (10-15). By
using computed tomography (CT) images without magnification, distortion, and superposition, 3D
* Department of Dental Service, Etimesgut Military Hospital
** Department of Orthodontics, Center of Dental Sciences of Gulhane Military
Medical Academy
*** Department of Orthodontics, Dental Service of GATA Haydarpasa Teaching
Hospital
Reprint request: Şeniz Karaçay, Department of Orthodontics, Dental Service
of GATA Haydarpasa Teaching Hospital, Tıbbiye Caddesi, Üsküdar, İstanbul
E-mail: snzkaracay@gmail.com
Date submitted: October 22, 2010 • Date accepted: December 23, 201044 • March 2011 • Gulhane Med J Yıldırım et al.
reconstruction, segmentation, and simulation can
be done by means of computer programs (2,4,11).
However, there are no available standard norm values regarding 3D images since it is a new developing
technique (16). Although different 3D analysis methods have been recommended, interpreting a measurement or even more developing a treatment plan according to these analyses is rather difficult since the
exact standard values are still unknown (7,17).
While developing a new method that can be used
on 3D images the answers of the following questions are important: Can conventional cephalometry
measurements be carried out on 3D images? Are the
norm values used in conventional cephalometry in
accordance with 3D cephalometry? Can orthodontic
treatment planning be done on 3D images with merely the conventional cephalometric analyses made
on lateral cephalograms? When the answers of these
questions are found out, 3D cephalometry would be
more accurately and frequently used in orthodontics.
In the present study, it was aimed to compare the
three dimensional (3D) craniofacial measurements
with conventional two dimensional cephalometric
measurements in patients with skeletal Class III malocclusion. The principal goal was to search if conventional cephalometry could be used on 3D images
of the facial structures.
Material and Methods
In this retrospective study, pretreatment lateral cephalometric radiographs and CT images of 44 patients
with skeletal Class III (ANB<0°) malocclusion were
used. Thirty one of the selected patients were male
with a mean age of 21.6 years, and 13 of them were
females with a mean age of 20.4 years.
Lateral cephalometric radiographs were taken with
Odontorama P.C. radiography instrument (Trophy
Radiologie, France) by using 18x24 cm sized x-ray
films. The distance between patient and x-ray source
was 150 cm, while it was 12 cm between the film cartridge and the patient. These values were standard for
all radiographs. Computed tomography images were
obtained by using Philips MX 8000 IDT Multislice
CT System V 2.5 (Philips Medical Systems, The
Netherlands) instrument, at a dose of 120 kV and 100
mA, with a section thickness of 1 mm. These images were registered to compact disc environment in
DICOM format. While conventional cephalometric
measurements were performed by tracing the radiographs with a 0.3 mm pencil, 3D cephalometric analyses were made by converting tomography data to 3D
via Mimics®
v12.01 (Materialise, Leuven, Belgium)
software.
In 3D images, segmentation of soft tissue, vertebra, mandible, and cranium were done separately so
as to mark the cephalometric points more easily.
Cephalometric analysis program of Mimics®
simulation module was used for cephalometric measurements. After the anatomic points were marked on
3D model and controlled via sagittal, coronal and
axial reformat sections (Figure 1), the program gave
the results of cephalometric analysis in two different
numerical value forms as 3D and 2D (Figure 2). The
average values of the measurements belonging to the
right and left structures were used.
Figure 1. A,B. Marking cephalometric points on Mimics®
program
Figure 2. Cephalometric analysis results on Mimics®
program
B
AVolume 53 • Issue 1 Comparison of cephalometric measurements • 45
Table I. Landmarks used in the study
Landmark Abbreviation Description
1. Nasion N The most anterior point of the nasofrontal suture on midsagittal plane
2. Sella S The center of the hypophyseal fossa (sella turcica)
3. Orbitale Or The lowest point on the lower magrin of each orbit
4. Porion Po The upper magrin of the porus acusticus externus
5. Anterior nasal spine ANS The most anterior point of the maxilla on midsagittal plane
6. Posterior nasal spine PNS The most posterior point at the sagittal plane on the bony hard palate
7. Point A A Deepest point on midsagittal plane between ANS and prosthion
8. Point B B The deepest midline point on the mandible between infradentale and pogonion
9. Pogonion Pog The most anterior pointon the mandible in the mildine
10. Menton Me Most inferior point on the symphysis of the mandible in the median plane
11. Gnathion Gn The most anterior-inferior point of the bony chin
12. Gonion Go A posterio-inferior point on the ramus
13. Condylion Co The most posterior superior point on the condyle of the mandible
14. Incision superius Is1u Tip of incisal edge of anteriormost upper incisor
15. Upper incisor apex Ap1u The root apex of the most prominent upper incisor
16. Incision inferius Is1l Tip of incisal edge of anteriormost lower incisor
17. Lower incisor apex Ap1l The root apex of the most prominent lower incisor
18. Occlusal 1 Occ1 Upper and lower first molar occlusal contact point
19. Occlusal 2 Occ2 Overbite midpoint
20.Labrale superius Ls The most anterior point on the magrin of the upper membranous lip
21. Labrale inferius Li The most anterior point on the magrin of the lower membranous lip
22. Pronasale Ns The most anterior point on the midsagittal profile of the nose
23. Soft tissue pogonion Pog’ The most anterior point on the soft tissue chin in the midsagittal plane
24. Soft tissue nasion N’ The deepest point on the concavity overlying the area of the frontonasal suture
In this study, 24 cephalometric points were used
(Table I and Figure 3). The selection of the points was
made by considering the frequency of the usage in
orthodontics. The points of intersection or superposition were not selected since they were hardly deFigure 3. Determined anatomic landmarks
tected on 3D images. 14 angular (Figure 4) and 18
linear (Figure 5) measurements were performed by
conventional method, MIMICS 3D, and MIMICS 2D
Figure 4. Selected angular measurements 1.SNA, 2.SNB, 3.ANB,
4.Y-Axis, 5.SN/ANS-PNS, 6.SN/Occ, 7.SN/Go-Gn, 8.ANS-PNS/
Go-Gn, 9.U1/SN, 10.U1/NA, 11.L1/Go-Gn, 12.L1/NB, 13.U1/L1,
14.H-Angle46 • March 2011 • Gulhane Med J Yıldırım et al.
(p=0.012), SN/Occ (p<0.001), SN/Go-Gn (p=0.002),
ANS-PNS/Go-Gn (p=0.001), U1/SN (p=0.016), U1/NA
(p=0.001), L1/NB (p=0.020) and H-Angle (p=0.005) revealed statistically significant differences in the comparison of Group II and Group III. All of the parameters, except U1/SN, were higher in Group II (Table II).
In the examination of the linear measurements, significant differences were detected in S-N (p<0.001),
Go-Me (p<0.001), Co-A (p<0.001), Co-Gn (p=0.003),
N-Me (p<0.001), N-ANS (p<0.001), ANS-Me (p<0.001),
S-Go (p<0.001), U1-NA (p<0.001), L1-NB (p=0.004),
LL-E (p=0.028), UL-Length (p<0.001) and S-Go/N-Me
ratio (p<0.001) between Group I and Group II. All of
the linear measurements, except Go-Me, Co-A, S-Go
and S-Go/N-Me, were higher in Group I according
to Group II. The comparison of Group I and Group
III revealed significant differences in S-N (p<0.001),
Go-Me (p<0.001), Co-A (p<0.001), Co-Gn (p<0.001),
N-Me (p<0.001), N-ANS (p<0.001), ANS-Me (p<0.001),
S-Go (p<0.001), U1-NA (p<0.001), L1-NB (p<0.001),
LL-E (p=0.031), UL-Length (p<0.001). All of the linear measurement values in Group I were higher
than the values in Group III. In the comparison of
Group II and Group III, NPerp-A (p=0.011), NPerpPog (p=0.005), S-N (p<0.001), Go-Me (p<0.001), Co-A
(p<0.001), Co-Gn (p<0.001), N-ANS (p<0.001), S-Go
(p<0.001), S-Go/N-Me (p<0.001), U1-NA (p<0.001),
L1-NB (p<0.001), Overjet (p<0.001), UL-E (p<0.001),
LL-E (p=0.001), UL-Length (p=0.005) showed statistically significant differences. All linear measurement
values, in Group II were greater than the values in
Group III. In other words, the findings of our study
revealed that the linear measurements showed the
highest values in Mimics 3D cephalometry and the
lowest values in Mimics 2D cephalometry.
Discussion
Cephalometric analyses have been frequently used
as the primary diagnosis tool in orthodontics for the
assessment of craniofacial structures. However, despite
their advantages such as low cost, low radiation dose,
and high reproducibility, they still have some shortcomings because of the superimposition of structures of the left and right side of the skull, the unequal
enlargement ratios of the left and right side, and the
possible distortion of the mid-facial structures (18,19).
Recently, 3D cephalometry obtained from CT scans
has been developed as an alternative to cephalometric
analysis. In this technique, the linear and angular measurements are made directly on 3D surfaces (16,20).
In the first studies, cranium was monitorized with
CT and only axial section data was examined without
applying 3D reconstruction and the researchers evaFigure 5. Selected linear measurements 1.NPerp-A, 2.NPerp-Pog,
3.S-N, 4.Go-Me, 5.Co-A, 6.Co-Gn, 7.N-Me, 8.N-ANS, 9.ANSMe, 10.S-Go, 11.S-Go/N-Me, 12.U1-NA, 13.L1-NB, 14.Overjet,
15.Overbite, 16.UL-E(Ls to E- Line), 17.LL-E(Li to E-Line), 18.ULLenght(ANS-Ls)
and the results were categorized as Group I, Group II,
and Group III, respectively (Table II).
Measurements carried out in three different ways
were evaluated by applying repeated measures analysis of variance. In order to determine probable differences among 3 experimental groups compared,
Wilks’ Lambda statistical analysis method was used.
When a difference was observed, Bonferroni posthoc test was used in order to detect the source of difference. In the evaluation of the parameters which
were not in accordance with normal distribution,
Friedman non-parametric repetitive measurement
analysis was applied. When difference was observed
between the groups, Bonferroni corrected Wilcoxon
non-parametric two dependent sample test was used
to determine the differences. For all statistical analyses and calculations, MS-Excel and SPSS for Win. Ver.
15.00 (SPSS Inc. Chicago, IL., USA) packaged software
were used. The significance level was set at p< 0.05.
Results
Evaluation of the angular measurements revealed
statistically significant differences in SNA (p<0.001),
SNB (p=0.10), ANB (p=0.001), ANS-PNS/Go-Gn
(p=0.023) and U1/L1 (p=0.046) values between Group
I and Group II and all of the parameters were higher
in Group II. Comparison of Group I with Group III
showed statistically significant differences in SNA
(p<0.001), SNB (p=0.07), ANB (p=0.004), L1/Go-Gn
(p=0.018) and U1/L1 (p=0.021). All of the parameters,
except L1/Go-Gn, were higher in Group III. Y-Axis Volume 53 • Issue 1 Comparison of cephalometric measurements • 47
luated these images with a physical anthropological
point of view without considering the orthodontic
points (21,22). In some previous studies, cephalometric radiographs and CT images were compared with
physical measurements, and the results revealed some
important differences between conventional cephalometric measurements and physical measurements of
cranium (5,23), whereas the measurements of 3D CT
images were closer to physical measurements (24-26).
Lopes et al. used 28 dry skulls and 3D CT images to
examine the accuracy and sensitivity of angular measurements and stated that there was no difference
between the two groups (16). Similarly, Chidiac et al.
used 13 skulls to compare the conventional cephalometric measurements and the measurements carried
out on CT images with each other and with physical
Table II. Mean and standard deviation values, inter-group distributions of repeated measurements analysis of variance and Bonferroni
post-hoc test
Group I (Conventional)
mean ± std dev.
Group II (MIMICS 3D)
mean ± std dev.
Group III (MIMICS 2D)
mean ± std dev.
test I-II I-III II-III
Skeletal
Skeletal angular
SNA 78.455±4.043 79.626±4.186 79.625±4.173 *** *** *** ns
SNB 83.705±5.165 84.455±5.331 84.490±5.335 ** * ** ns
ANB -5.250±3.498 -4.830±3.499 -4.865±3.471 * ** ** ns
Y-Aksis 57.750±5.248 57.685±5.838 57.562±5.772 ** ns ns *
SN/ANS-PNS 9.068±4.049 8.616±4.345 8.603±4.340 * ns ns ns
SN/Occ 14.955±6.548 16.346±9.020 15.929±9.124 ** ns ns ***
SN/Go-Gn 34.455±7.125 34.371±7.124 34.160±7.266 ** ns ns **
ANS-PNS/Go-Gn 25.205±6.465 25.959±6.778 25.680±6.889 *** * ns **
Skeletal linear
NPerp-A -3.364±3.314 -3.038±3.201 -3.033±3.196 ** ns ns *
NPerp-Pog 8.205±8.733 9.310±8.744 9.294±8.729 ** ns ns *
SN 74.386±4.701 67.242±4.079 67.098±4.027 *** *** *** ***
Go-Me 83.545± 6.407 89.575±5.820 75.281±6.003 *** *** *** ***
Co-A 90.636±7.038 98.080±6.358 83.420±6.083 *** *** *** ***
Co-Gn 138.409±10.228 136.330±8.022 126.153±7.945 *** ** *** ***
N-Me 138.795±11.099 126.047±9.893 125.949±9.838 *** *** *** ns
N-ANS 59.386±4.794 53.637±4.201 53.596±4.220 *** *** *** ***
ANS-Me 79.955±8.441 72.944±7.633 72.788±7.609 *** *** *** ns
S-Go 89.386±8.269 93.613±6.992 80.060±7.082 *** *** *** ***
S-Go/ N-Me 0.646±0.054 0.745±0.049 0.637±0.050 *** *** ns ***
Dental
Dental angular
U1/SN 106.614±7.851 106.349±7.851 106.391±7.851 * ns ns *
U1/NA 27.364±6.567 26.854±6.681 26.745±6.676 ** ns ns **
L1/Go-Gn 78.227±7.554 78.540±6.761 76.247±8.112 *** ns * ***
L1/NB 15.841±6.038 15.416±6.529 15.032±6.773 * ns ns *
U1/L1 140.977±10.936 142.888±11.867 143.108±12.100 * * * ns
Dental linear
U1-NA 5.977±2.445 3.693±2.253 2.179±2.353 *** *** *** ***
L1-NB 3.080±2.023 2.345±2.290 1.186±1.521 *** ** *** ***
Overjet -4.261±3.918 -4.005±3.657 -3.898±3.586 ** ns ns ***
Overbite 0.045±3.906 0.252±4.354 0.252±4.347 ns ns ns ns
Soft tissue
Soft tissue angular
H-Angle (N’Pog’Ls) 3.159±5.048 3.158±5.296 2.773±4.737 * ns ns **
Soft tissue linear
UL-E (Ls-E Line) -10.182±3.105 -10.273±2.668 -10.253±2.663 *** ns ns ***
LL-E (Li-E Line) -3.409±3.029 -3.853±2.716 -3.846±2.713 *** * * **
UL-Length (ANS-Ls) 25.727±3.216 21.956±2.837 21.709±3.047 *** *** *** **
*p<0.05, **p<0.01, ***p<0.001, ns: non significant48 • March 2011 • Gulhane Med J Yıldırım et al.
measurements (25). The authors reported significant
differences at linear measurement values between conventional cephalometry and CT images. Nevertheless,
there were no differences at angular measurements.
Togashi et al. investigated whether the position of
the head affected the accuracy of the linear measurements performed on 3D CT images (20). The investigators obtained CT images of the cranium with
thicknesses of 1 mm, 3 mm, 5 mm, and 7 mm at different head positions. They concluded that the measurements made on 3D CT images were independent
from the head position but the increase in the section
thickness might cause failures in some linear measurements. Similarly, Kitaura et al. (7), Cavalcanti et al.
(24), and Park et al. (15) showed that the axial section thicknesses of CT images affected the quality of
3D images. Kitaura et al reported that, the 3D images
obtained with section thickness less than 3 mm were
almost without failure when compared to actual values (7). On the other hand, it was reported that, in
contrast to CT images, the inappropriate head position affected the accuracy of the linear and angular
measurements made on lateral cephalograms (27,28).
In the light of these findings, the CT images were
obtained with an axial section thickness of 1 mm in
our study. Although some authors reported that the
head position did not affect the CT images (7,20), we
obtained the CT images with Gantry and Tilt values
of 0° (Frankfurt Horizontal plane was perpendicular
to ground plane without head rotation), taking into
consideration the suggestions of the computer program that was used in our study.
Lateral cephalograms are 2D projection images of
3D objects. For this reason, anatomic structures are
usually subjected to either vertical or horizontal displacement depending on the distance between the objects and the radiograph. It is notified that, magnification is observed in most of the craniofacial structures, with rates varying from approximately 0% in
structures at the side close to the film and on central
ray, and to 24% in structures which are 60 mm or
more far from ear sticks (5-9). Because of that, the
use of a constant magnification correction for every
measurement value carried out on 2D conventional
cephalograms could lead to some mistakes. The results of our study revealed that the differences in the
magnifications observed in 2D conventional cephalograms and 3D images were statistically significant
in linear measurements. The differences were varying
between 1.52% and 38.21% and no constant rate was
observed. In the light of this finding, it can be concluded that a standard magnification correction for
every measurement could create inaccurate results.
The Gonion (Go) and Condilion (Co) points are
far from mid-sagittal plane and belong to bilateral
structures. Kragskov et al. who compared the conventional 2D cephalometric measurements and
the measurements applied on 3D CT images stated
that in oblique measurements of bilateral structures
and points on the mid-sagittal plane, the results of
conventional cephalometry values were lower than
physical and 3D measurement values (29). The magnification in conventional cephalograms, made the
values of Go-Me, Co-A and S-Go closer to the 3D measurement. However, the values of Group I were still
lower than those of Group II in our study. Although
it has the same characteristics, Co-Gn measurement
was higher in conventional cephalometry (Group I),
when compared to the 3D cephalometric measurements (Group II). Gnathion (Gn) point is one of the
furthest points from the central ray, and this may be
the reason of the higher magnification observed in
the conventional cephalometry. These findings are
in accordance with the findings of Adams et al (5)
and Kragskov et al (29).
Kumar et al. compared the images obtained via
Cone Beam CT with conventional cephalograms and
stated that the differences of ± 2 degrees between
angular measurements and of ±2 mm between linear measurements were clinically insignificant (30).
Cavalcanti et al. (24), Periago et al. (31), and Jamali
et al. (32) also reported similar results. In our study,
the differences between the groups were lower than 2
degrees in the angular measurements. Additionally, it
was lower than 2 mm in the linear measurements of
Nperp-A, Nperp-Pog, Overjet, Overbite, L1-NB, UL-E
and LL-E. In the light of findings reported by the above authors, it can be stated that although some of
these parameters were statistically significant; these
small differences may be clinically ignored.
In the light of our findings the followings can be
concluded: 3D visualization is an ideal visualization
technique for orthodontics since it enables more sensitive measurement and planning. In conventional
cephalometry, and computer aided 3D and 2D cephalometry, the angular measurements were consistent with each other. In linear measurements, except
the measurement values of of Nperp-A, Nperp-Pog,
Overjet, Overbite, L1-NB, UL-E and LL-E, significant
differences were observed. No current norm values
are available for 3D cephalometry. Until a database
is developed for 3D cephalometry, the conventional
cephalometric norms may be used for the angular measurements. In conventional cephalometry, standard
magnification correction for every measurement may
create inaccurate results