(Doi 10.1043 - 0003-3219 (2006) 076 (0388 - VOPRFM) 2.0.CO 2)
(Doi 10.1043 - 0003-3219 (2006) 076 (0388 - VOPRFM) 2.0.CO 2)
(Doi 10.1043 - 0003-3219 (2006) 076 (0388 - VOPRFM) 2.0.CO 2)
Abstract: The objective of this study is to evaluate the accuracy of panoramic radiographs for
diagnosing vertical asymmetry of the posterior mandible. The first part of the study used a model
to evaluate the reproducibility of this particular panoramic machine. The tube traverse did not
significantly affect the linear measurements, but the side of the machine where the structure was
located produced an average of 2.1% variation in the total height of the mandible. In the second
part of the study, the left-right (%) differences were measured on the panoramics and the lami-
nographs of five skulls with lead markers. These differences were compared with the percent
difference measured directly on the skull and with each other and suggested that the laminograph
could be used as the ‘‘gold standard’’ for measuring posterior vertical mandibular asymmetry. The
third part of the study evaluated a patient sample (N 5 42). The correlation coefficient between
the two types of films was determined to be 0.92 for total height and 0.39 for condyle height.
Using the 6% cutoff reported in the literature, the sensitivity of the panoramics to diagnose asym-
metry for the total height was determined to be 0.62 and the specificity 1.0. This study suggests
that the panoramic radiograph can be used to evaluate vertical posterior mandibular asymmetry,
but there will be some underdiagnosis. (Angle Orthod 2006;76:388–393.)
Key Words: Mandibular asymmetry; Panoramic; Laminograph; TMJ
FIGURE 2. Two films taken on the panoramic machine, one using the phantom as the right and the other as the left half of the mandible.
TABLE 1. Multiple Regression Analysis Comparing Actual vs Ra- TABLE 2. Multiple Regression Analysis Comparing Panorex With
diographic Measurements on the Phantom. All the Correlation Esti- Laminagraphic Measurements in the Patient Sample
mates Were Significant (P , .05)
Laminagraph
R Difference Laminagraph
(Yes) Difference (No)
Panorex—actual Total height 0.95
Condylar height 0.95 Panoramic difference (Yes) 8 0
Laminagraph—actual Total height 0.99 Panoramic difference (No) 5 29
Condylar height 0.99
Panoramic—laminograph Total height 0.95
Condylar height 0.97 TABLE 3. In the Patient Sample: Using the Laminagraph as the
Gold Standard and the 6% Cutoff the Sensitivity and Specificity of
the Two Diagnostic Tests Were 0.62 and 1.0, Respectively
RESULTS Panorex vs
Part 1: Plexiglas phantom Laminagraph Total Height Condylar Height
R 0.92 0.39
The traverse of the machine was tested using the
phantom. Two series of films were performed, one us-
ing the phantom as the right half and one as the left
between the panoramic radiograph and laminagraphic
half of the mandible. When the model was positioned
measurements.
on the left side, the mean condylar height measure-
Using the laminagraph as the gold standard and a
ment and the mean total height measurement was
6% cutoff17 as an indication of potential asymmetry, a
measured for the L-R and the R-L traverse. Paired t-
2 3 2 table was constructed (Table 3) to evaluate the
test (two-tailed) suggested that these differences were
sensitivity and specificity of the two tests. The sensi-
not significant (P . .05). The same was found when
tivity and specificity of the diagnostic tests were 0.62
the model was positioned in the right side as the
and 1.0, respectively. Therefore, a panoramic radio-
paired t-test (two-tailed) for analysis. These differenc-
graph will not overestimate the asymmetry in this pa-
es were not significant (P . .05).
tient population if we use 6% as our cutoff for true
However, when the R-L and L-R traverses were
asymmetry. There were no false-positive examinations
pooled together and the means of the condylar and
with the panoramic radiograph.
the total height for the right side vs the left were sub-
jected to the paired t-test (two-tailed) for statistical
DISCUSSION
analysis, the differences were significant (P , .05).
The clinically significant difference for total height was The purpose of this study was to evaluate the po-
found to be a 2.1% difference between the two sides. tential of panoramic radiographs to determine vertical
asymmetry between the right and left side in the man-
Part 2: Skulls dible. Measurements of the total height on the phan-
tom were compared with the measurements on the ra-
The multiple regression analysis for the five skulls
diographs. The consistency of the measurements for
demonstrated high correlation between the measure-
those structures was determined comparing the differ-
ments on the skulls and the radiographs, which was
ent tube traverses and the phantom placed on the left
always more than 0.95 (Table 1). The regression for
vs the right side. This produced good reproducibility
the comparison of the actual measurements on the
for the panoramic machine for repeated positioning
skulls with those on the laminagraphs was 0.99 and
and exposures with different tube travel. The tube trav-
supported the use of the laminagraphs as the gold
el appeared to be negligible on this machine. How-
standard in the patient study.
ever, the side that the phantom was positioned was
not negligible. There was a 1.45-mm difference be-
Part 3: Patient sample
tween right and left side for total height, which was
The asymmetry measured on the panoramic radio- calculated to be a 2.1% difference between the two
graph and that measured on the laminagraphs taken sides. Thus, it appears that there is an average inher-
from the patients had a correlation of 0.92 for total ent error in the machine of 2.1% for total height.
height and 0.39 for condyle height (Table 2). The The skull specimens allowed a direct measure of the
mean disagreement (the mean of the difference per- object being radiographed. The most accurate method
cent of measured asymmetry) between the panoramic to measure the skulls is to use lead markers because
and the laminagraph for the total height asymmetry they can easily be identified on the radiographs. The
was 2.23% and for condyle height was 11.9%. There panoramic and the laminagraph were both good, but
was never more than 6% disagreement for total height the laminagraph was better when compared with the
actual specimen (skull), and the correlation coefficient metry in patients after the 6% cutoff factor and using
was equal (R 5 0.99) for total height and condylar the laminographs as the gold standard. This would in-
height. The validity of the laminographs as the gold dicate that when using 6% as a cutoff for asymmetry,
standard was investigated. The results of the skull por- the panoramic radiograph would underdiagnose asym-
tion support the laminograph as the radiographic gold metry compared with the laminographic radiograph.
standard because we cannot compare the panoramic The specificity of the panoramic radiograph was 1.0
with the actual subject in live patients. (no false positive). This meant that none of the pano-
Measurements from the panoramic films were com- ramic radiographs indicated greater than 6% asym-
pared with the laminographic films for the worst-case metry if the laminograph radiographs indicated less
scenario for potential error in the measurements. The than 6% asymmetry. Therefore, when using the pan-
inherent error in the machine and operator error must oramic radiograph to detect mandibular total ramal
be considered. Both intra- and interoperator error may height differences, one can confidently state that there
be present in positioning and exposing the radio- is a true asymmetry if indicated on the panoramic us-
graphs, as well as intraoperator error in the tracing and ing the 6% cutoff. If one accepts the laminograph ra-
the measurements. diographs as the gold standard, the panoramic radio-
In the patient part of the study, there was a very low graph is reasonable for detecting asymmetries of total
correlation between panoramic and laminograph ra- ramal height.
diographs for the condyle height asymmetry (R 5
0.39), suggesting that the panoramic radiograph CONCLUSIONS
should be questioned for this measurement. This is
• Condylar height was unreliable when determining
probably because of the small dimension of the mea-
asymmetry from the panoramic radiograph.
surement and operator error in tracing and identifica-
• Total ramal height could be used in determining
tion of landmarks. This agrees with the findings of Turp
asymmetry from the panoramic radiograph, but it
et al,29 who evaluated condyle height and found the
must be realized that it would generally underdiag-
correlation to be very low when using Bezuur’s 6%
nose asymmetry.
cutoff for asymmetry. The total height measurement
• Some patients with a less than 6% difference be-
was more encouraging with R 5 0.92. This supports
tween the left and right sides might not be diagnosed
the use of the panoramic radiograph for evaluation of
with panoramic radiograph.
total ramal height asymmetry.20
In the patient sample, the mean disagreement be- REFERENCES
tween the panoramics and the laminographs for total
height was 2.23% and compares well with the first part 1. Gidarakou IK, Tallents RH, Stein S, Kyrkanides S, Moss
ME. Comparison of skeletal and dental morphology in
of the study that found the inherent machine error of asymptomatic volunteers and symptomatic patients with
2.1%. However, it cannot be deduced that these two unilateral displacement with reduction. Angle Orthod. 2004;
numbers represent the same error. The maximum dis- 74:212–219.
agreement for the total ramal height was 5.38%, sim- 2. Gidarakou IK, Tallents RH, Stein S, Kyrkanides S, Moss
ilar to the 6% cutoff that Habets et al20 calculated as ME. Comparison of skeletal and dental morphology in
asymptomatic volunteers and symptomatic patients with
reasonable for the panoramic radiograph. unilateral displacement without reduction. Angle Orthod.
In investigating morphologic effects of temporoman- 2003;73:121–127.
dibular dysfunction, Habets et al17,20 and Bezuur et al19 3. Gidarakou IK, Tallents RH, Stein S, Kyrkanides S, Moss
found that vertical differences between the left and ME. Comparison of skeletal and dental morphology in
right sides were less than 6% if positions were altered asymptomatic volunteers and symptomatic patients with bi-
lateral degenerative joint disease. Angle Orthod. 2003;73:
less than 10 mm from their original centered position 71–78.
(using an experimental model resembling a human 4. Bezuur JN, Habets LL, Hansson TL. The recognition of
mandible). They concluded that observed condylar craniomandibular disorders; condylar symmetry in relation
asymmetry within a 6% difference might, therefore, be to myogenous and arthrogenous origin of pain. J Oral Re-
because of technical failures. They also reported that habil. 1989;16:257–260.
5. Qadan S, Macher DJ, Tallents RH, Kyrkanides S, Moss ME.
a significant difference between controls and patients The effect of surgically induced anterior disc displacement
treated for craniomandibular disorders was found in of the temporomandibular joint on the midface and cranial
condylar height symmetry. We decided from previous base. Clin Orthod Res. 1999;2:124–132.
observations in this study that the use of a 6% cutoff 6. Hatala MP, Macher DJ, Tallents RH, Spoon M, Subtelny JD,
proposed by Bezuur et al19 was acceptable for com- Kyrkanides S. Effect of a surgically created disk displace-
ment on mandibular symmetry in the growing rabbit. Oral
paring right with left differences. Surg Oral Med Oral Pathol Oral Radiol Endod. 1996;82:
The panoramic radiograph was found to have a sen- 625–633.
sitivity of 0.62 (high false negative) in detecting asym- 7. Tallents RH, Guay JA, Katzberg RW, Murphy W, Proskin H.
Angular and linear comparisons with unilateral mandibular craniomandibular disorders; condylar symmetry in relation
asymmetry. J Craniomandib Disord. 1991;5:135–142. to myogenous and arthrogenous origin of pain. J Oral Re-
8. Schellhas KP, Piper MA, Omlie MR. Facial skeleton remod- habil. 1989;16:257–260.
eling due to temporomandibular joint degeneration: an im- 20. Habets LL, Bezuur JN, van Ooij CP, Hansson TL. The or-
aging study of 100 patients. Am J Neuroradiol. 1990;11: thopantomogram, an aid in diagnosis of temporomandibular
541–551. joint problems. I. The factor of vertical magnification. J Oral
9. Schellhas KP. Unstable occlusion and temporomandibular Rehabil. 1987;14:475–480.
joint disease. J Clin Orthod. 1989;23:332–337. 21. Larheim TA, Johannessen S, Tveito L. Abnormalities of the
10. Schellhas KP, Keck RJ. Disorders of skeletal occlusion and temporomandibular joint in adults with rheumatic disease. A
temporomandibular joint disease. Northwest Dent. 1989;68: comparison of panoramic, transcranial and transpharyngeal
35–39, 42. radiography with tomography. Dentomaxillofac Radiol.
11. Trpkova B, Major P, Nebbe B, Prasad N. Craniofacial asym- 1988;17:109–113.
metry and temporomandibular joint internal derangement in 22. Tronje G, Welander U, McDavid WD, Morris CR. Image dis-
female adolescents: A posteroanterior cephalometric study. tortion in rotational panoramic radiography. IV. Object mor-
Angle Orthod. 2000;70:81–88. phology; outer contours. Acta Radiol Diagn (Stockholm).
12. Larheim TA. Comparison between three radiographic tech- 1981;22:689–696.
niques for examination of the temporomandibular joints in 23. Welander U. A mathematical model of narrow beam rotation
juvenile rheumatoid arthritis. Acta Radiol Diagn (Stock- methods. Acta Radiol Diagn (Stockholm). 1974;15:305–
holm). 1981;22:195–201. 317.
13. Larheim TA, Hoyeraal HM, Stabrun AE, Haanaes HR. The 24. Larheim TA, Svanaes DB. Reproducibility of rotational pan-
temporomandibular joint in juvenile rheumatoid arthritis. Ra- oramic radiography: mandibular linear dimensions and an-
diographic changes related to clinical and laboratory param- gles. Am J Orthod Dentofacial Orthop. 1986;90:45–51.
25. Ramstad T, Hensten-Pettersen O, Mohn E, Ibrahim SI. A
eters in 100 children. Scand J Rheumatol. 1982;11:5–12.
methodological study of errors in vertical measurements of
14. Ronning O, Valiaho ML. Involvement of the facial skeleton
edentulous ridge height on orthopantomographic radio-
in juvenile rheumatoid arthritis. Ann Radiol (Paris). 1975;18:
grams. J Oral Rehabil. 1978;5:403–412.
347–353.
26. Ericsson S, Lundberg M. Alterations in the temporomandib-
15. Ronning O, Valiaho ML. Progress of mandibular condyle
ular joint at various stages of rheumatoid arthritis. Acta
lesions in juvenile rheumatoid arthritis. Proc Finn Dent Soc.
Rheumatol Scand. 1967;13:257–274.
1981;77:151–157.
27. Larheim TA, Svanaes DB, Johannessen S. Reproducibility
16. Athanasiou AE, Melsen B, Mavreas D, Kimmel FP. Sto- of radiographs with the Orthopantomograph 5: Tooth length
matognathic function of patients who seek orthognathic sur- assessment. Oral Surg Oral Med Oral Pathol. 1984;58:736–
gery to correct dentofacial deformities. Int J Adult Orthod 741.
Orthognath Surg. 1989;4:239–254. 28. Kjellberg H, Ekestubbe A, Kiliaridis S, Thilander B. Condylar
17. Habets LL, Bezuur JN, Naeiji M, Hansson TL. The Ortho- height on panoramic radiographs. A methodologic study
pantomogram, an aid in diagnosis of temporomandibular with a clinical application. Acta Odontol Scand. 1994;52:43–
joint problems. II. The vertical symmetry. J Oral Rehabil. 50.
1988;15:465–471. 29. Turp JC, Vach W, Strub JR, Harbich K, Alt KW. The rec-
18. Wabeke KB, Spruijt RJ, Habets LL. Spatial and morphologic ognition of mandibular asymmetries in the panoramic to-
aspects of temporomandibular joints with sounds. J Oral mogram. An example of the need for judging the value of a
Rehabil. 1995;22:21–27. diagnostic procedure. Schweiz Monatsschr Zahnmed. 1995;
19. Bezuur JN, Habets LL, Hansson TL. The recognition of 105:755–759.