Sunteți pe pagina 1din 5

i n d i a n j o u r n a l o f d e n t i s t r y 4 ( 2 0 1 3 ) 8 3 e8 7

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/ijd

Original Article

Comparative study of the effect of direct and indirect digital


radiography on the assessment of proximal caries
Farhad Aghmasheh a,*, Roghaye Bardal b, Zohreh Reihani c, Mohammad Ali Moghaddam d,
Sedighe Rahro Taban e, Farnoosh Fallahzadeh f, Anahid Ahmadi c
a

Department of Oral and Maxillofacial Radiology, Zanjan School of Dentistry, Safa St., Zanjan, Iran
Department of Oral and Maxillofacial Radiology, Qazvin School of Dentistry, Iran
c
Zanjan University of Medical Science, Iran
d
Department of Restorative Dentistry, Zanjan School of Dentistry, Iran
e
Department of Oral Pathology, Qazvin School of Dentistry, Iran
f
Department of Restorative Dentistry, Qazvin School of Dentistry, Iran
b

article info

abstract

Article history:

Introduction: Clinical assessment of proximal tooth contacts leads to unacceptable false

Received 15 August 2012

negative results especially in cases of tight proximal contacts where precise clinical

Accepted 11 October 2012

assessment is not possible. Therefore radiographic evaluation seems imperative for


proximal caries detection.

Keywords:

Materials and methods: 70 posterior permanent teeth were divided into blocks of 5 teeth.

Bitewing radiography

Digital and conventional Bitewing radiographs with D-speed and E-speed films were taken

Digital dental radiography

from samples. Conventional radiographs were digitalized with a scanner whereas direct

Dental caries

digital images were saved in either standard or filtered mode. Presence of proximal caries
was evaluated by three observers and samples were scored accordingly. As golden standard, teeth were then sectioned horizontally for histological evaluation of caries depth.
Results: Diagnostic specificity of each radiography method was evaluated with sub curve
analysis of ROC (Az). Spearman test was used for inter- and intra-observers agreement. No
significant difference was reported between the sensitivity and specificity of the four
studied methods (P > 0.05). The digital filtered image had the highest sensitivity. The
highest specificity was attributed to D-speed films.
Conclusion: Indirect digital images of scanned D-speed and E-speed films and direct standard and filtered digital images (CMOS receptor) have high specificity and there is no significant difference in the diagnosis of proximal caries between them.
2012 Indian Journal of Dentistry. All rights reserved.

1.

Introduction

Dental caries is the most common infectious disease among


human.1 Sole clinical assessment of proximal contacts leads

to unacceptable false negative results especially in cases of


tight proximal contacts where precise examining is not
possible.2 Hence both clinic and radiographic measures are
paramount for the detection of proximal caries.3 Suitable

* Corresponding author. Tel./fax: 98 241 4244001.


E-mail address: f.aghmasheh@gmail.com (F. Aghmasheh).
0975-962X/$ e see front matter 2012 Indian Journal of Dentistry. All rights reserved.
http://dx.doi.org/10.1016/j.ijd.2012.10.008

84

i n d i a n j o u r n a l o f d e n t i s t r y 4 ( 2 0 1 3 ) 8 3 e8 7

treatment planning is the major objective of production and


evaluation of radiographic images. In other words, the difference in treatment planning is a consequence of different
radiographic interpretation and similarly different diagnostic
abilities.2,4 Diagnosis of proximal caries with bitewing radiograph is a well-accepted and confirmed method.5,6 Both conventional and digital bitewing radiographies are evaluated in
this study. Conventional films are the most common receptors for intraoral radiography.7 Various film speeds are
introduced into the market. Nowadays reduction of radiation
dose is an unceasing concept in the production of different
intraoral films.8 Current study assessed D-speed and E-speed
films for any correlation between the higher speed of film and
reduction of diagnostic accuracy. Digital systems have many
advantages such as reduced exposure, elimination of film
processing, variable contrast and density and the use of
software.9e11 This system is however more expensive and has
so far yielded contrasted results regarding its diagnostic
ability.12,13 Use of digital receptor is not common yet, therefore conventional radiography is still used in most cases. So
long as the digital systems have not proved popular, digitalization of conventional images is required.14 Indirect digital
radiography was first introduced in 1970s. In this method,
conventional films were used as image receptors. A scanner or
camera then digitalized the image.15,16 The method is called
optically scanned digital radiography.15 No persistent
agreement have been reached in previous studies regarding
the effect of the image software alterations on the diagnostic
accuracy5,7,17 there have also been a limited number of studies
on the comparison of direct and indirect digital images .In
some studies no significant difference was reported between
these two methods7,18 while some other studies have proved
otherwise.19 The aim of this study was to compare the accuracy of indirect digital radiography images of D- and E-speed
films with standard and filtered direct digital radiography in
the proximal caries detection.

2.

Materials and methods

70 extracted posterior teeth (28 molars and 42 premolars) were


selected. Caries definition varied from a small discoloured
area to a real cavity but the crown and root integrity were not
altered in all cases. Teeth with restoration, buccal or lingual
caries were excluded. Samples were then preserved in 2%
hypochlorite for 20 min and kept in physiologic serum till their
sectioning.20 Teeth were divided into groups of 5 teeth, each
comprising of 2 molars and 3 premolars and mounted in
polyester blocks so that 3 premolars were placed in the middle
and 2 molars on the two sides to create natural tooth contacts.
All blocks and teeth were numbered. In order to simulate the
bitewing technique, two blocks were mounted in an articulator. XCP Rinn (Rinn, USA) was used to standardize the bitewing projections. A 12 mm thick acrylic plate was placed
between X-ray tube and tooth blocks to simulate the soft tissues. Conventional and digital intraoral radiographs were
taken from all blocks. In the conventional method, size 2
Kodak E-speed (NY-USA) and size 2 Kodak D-speed (NY-USA)
films were used. CMOS receptor (Schick, USA) was utilized in
the digital technique. A 60 kvp, 8 mA exposure parameters by

Planmeca Intra device, (Planmeca, Finland) with stable


contrast and density was used for all radiographs. Exposure
time was determined according to the Pilot study; 1 s for D
film, 0.63 s for E film and 0.32 s for CMOS receptor. The focal
spot to the image receptor distance was constantly set on
35 cm. Films were processed automatically by an automatic
processor (peripro, Hoop, USA) with fresh developing and
fixing solutions (Champion, England). Conventional radiographs were scanned and digitalized at 300 dpi (Microteck
scan Maker, 6900 XL, Taiwan, R.O.C). One of the two drafts of
each image was filtered by a radiologist through the software
of receptor (Schick, USA) with the Edge enhancement filter.
Samples were then numbered and a distal and mesial surface
was hypothetically attributed to them. Only the proximal
sides with contacts were evaluated by observers. A radiologist
adjusted the density before the 4 series of images were
observed by other observers consist of two radiologists and
one operative dentistry specialist. Images were shown in a 17
inch monitor LCD (Samsung Master NW, 1733) (Fig. 1). Observers were not allowed to manipulate images. Images were
re-evaluated in two weeks in order to verify the intra observer
agreement. Proximal sides were assigned as: probable caries
free/certain caries free/unsure/probable caries/definite caries.
In the last two cases, the extent of caries had to be determined
and its depth was scored from 1 to 4 as caries in the: external
half of the enamel, internal half of the enamel, external half of
the dentin, internal half of the dentin accordingly.5,17,19,21,22
Hence 112 proximal sides were examined. 4 images were
captured of each side and evaluated by observers. Mesial and
distal sides of teeth were marked with a nail polish after
radiographic evaluation. For the histologic assessment and its
comparison with radiologic evaluation, teeth were separately
mounted in acrylic blocks. On each block, the tooth number,
its mesial and distal sides were defined. Teeth were horizontally divided into 700 micron thick sections with a 200 micron
thick blade of mecatome (Mecatome, T201, Presi Co). Sections
were evaluated by a pathologist with an 8e16 magnification
Stereomicroscope (MEC-2). Suspicious cases were consulted
with another oral pathologist. The depth of caries scored as
0 sound surface, 1 caries in outer half of enamel, 2 caries
in inner half of enamel, 3 caries in outer half of dentin and
4 caries in inner half of dentin.5,17

3.

Results

The histological evaluation of samples as gold standard


showed that in 12 surfaces of 112, the deepest part of the
caries had been located on the root surface, so these surfaces
excluded from the study and finally 100 proximal surfaces
were analysed. The ROC (receiver analysing curve) was drawn
to determine caries detection accuracy based on sub curve
surface in all mentioned methods (Table 1). Based on results,
the highest accuracy belongs to digitized E films but the differences between methods are not statistically significant.
Specificity and sensitivity for all radiographic methods estimated based on observers results related to different caries
depths in contrast with the gold standard (Tables 2 and 3). The
highest specificity belongs to digitized D film images and the
highest sensitivity for filtered direct digital ones. Statistical

85

i n d i a n j o u r n a l o f d e n t i s t r y 4 ( 2 0 1 3 ) 8 3 e8 7

Fig. 1 e Four types of dental radiographic images of a same block. (A): Digitized E film image. (B): Digitized D film radiograph
of the block. (C): Image of the same block prepared by CMOS receptor. (D): Edge-enhanced CMOS receptor image of the same
teeth.

analysis shows no significant differences between specificity


and sensitivity of the methods (P > 0.05). The results also show
that sound surface detection percentage is higher in all
methods in contrast to detection of caries. Dentinal caries are
also better recognized than enamel ones, and detection of
caries in inner part of dentin is more accurate than the outer
part. Tables 4 and 5 show Spearman test results for inter- and
intra-observer agreements accordingly.

4.

Discussion

Current study was designed to compare the diagnostic


sensitivity of direct standard and filtered digital radiography
with indirect digital radiography of D- and E-speed films for
the detection of proximal caries. According to the results of

ROC curve analysis, despite high sensitivity there was no


significant difference between studied methods.
The sub curve area of ROC analysis, being an indicator of
accuracy, was reported between 0.767 and 0.809. The largest
area belonged to E film and the smallest to the digital D film.
Seneadza et al scanned D and E films and filtered one copy of
the digital images. Comparison of the diagnostic sensitivity of
different images didnt reveal significant difference.5 This
result is in accordance with the result of current study. Other
studies have revealed no significant difference of diagnostic
sensitivity between conventional radiography of E- and Dspeed films and digital radiography.5,20,23e26 Uprichard et al

Table 2 e Specificity and sensitivity evaluated for each


radiographic sensor type and observer. (The upper
number shows sensitivity percentage and the lower one
is specificity percentage).
Observer

Table 1 e Diagnostic accuracy according to sub ROC curve


area.
Sensor
type

Accuracy

Std. error

Upper
limit

Lower
limit

D film
E film
CMOS
Filtered
CMOS

0.767
0.809
0.778
0.787

0.030
0.028
0.032
0.032

0.825
0.863
0.840
0.850

0.708
0.755
0.715
0.724

1
2
3
Total

Sensor type
D film

E film

CMOS

Filtered CMOS

50
96.2
60.8
60.8
66
73.1
63.4
89

58
92.3
74.3
80.8
75.7
69.2
75
87.5

60.8
84.6
67.6
69.2
91.9
46.2
79.75
76.5

63.5
88.5
70.3
65.4
94.6
73.1
82.45
84.3

86

i n d i a n j o u r n a l o f d e n t i s t r y 4 ( 2 0 1 3 ) 8 3 e8 7

Table 3 e Sensitivity, specificity and true diagnosis


percentage for different methods according to difference
histological caries depth. (Upper numbers are specificity
or sensitivity and lowers are true diagnosis percent).
Histological depth
4

44.5
20.3
55.7
28.9
58.8
29.4
68.7
44.9

31.5
10.5
48.1
26.3
57.8
26.3
66.6
44.4

30.3
9.5
45
10
50
0
61.1
0

Observer 1 and 2
Observer 1 and 3
Observer 2 and 3

0
Specificity
89
e
87.5
e
76.5
e
84.3
e

E film
CMOS
Filtered CMOS

Table 4 e Intra observer agreement results based on


Spearman coefficient.
Observer

Radiographic method
D film

E film

CMOS

Filtered CMOS

0.82
0.78
0.75

0.79
0.78
0.72

0.66
0.67
0.59

0.71
0.71
0.70

D film

reported the diagnostic accuracy of E- and D-speed films to be


higher than digital radiography.27 Their result is in contrast to
current study which could be due to the speciality of the observers. In Uprichards study, five paediatric dentists observed
images. The experience of observers, observation condition
such as monitor, receptor and exposing condition is another
justification for this contrast. Comparison of the diagnostic
sensitivity of digital images with filtered images revealed no
significant difference in many studies.22,26,28 This is in accordance with current study. In some studies however direct
digital images increased the diagnostic accuracy.17,29 The
contrast of this result with that of current study could be
justified by the type of the filter used. Type of receptor, software facilities, number and speciality of observers are also
accountable. There was no significant difference in the
sensitivity and specificity of different studied methods
(P > 0.05). Rockencbach et al revealed similar results in their
own study.23 In the current study the highest specificity (89%)
and the least sensitivity (63.4%) was attributed to the digitalized D-speed film and highest sensitivity (82.45%) to the
filtered digital image. The lowest specificity was attributed to
digital CMOS. Clinical implication of a high specificity and low
sensitivity of a radiograph is a better diagnosis of sound surface compared to decayed surface. False negative results are
consequently more stated than false positive ones. Hence in
many cases of wrong diagnosis, surfaces are considered
sound and no preparation is done. Specificity was reported to
be higher compared to sensitivity in mentioned methods.
Accordingly, even in filtered digital images that have higher
sensitivity compared to other methods, diagnosis of sound
surfaces is more reliable than decayed ones. This result is in

1
2
3

Observer

Radiography

Sensitivity and true diagnosis


percent
75.7
45.3
84.8
47.5
86.8
56.6
86.8
56.6

Table 5 e Intra observer agreement results according to


Spearman coefficient.

Radiographic method
D film

E film

CMOS

Filtered CMOS

0.89
0.77
0.82

0.90
0.82
0.85

0.76
0.80
0.71

0.81
0.79
0.78

accordance with Eortens study.12 In all studied methods,


sensitivity and the number of correctly diagnosed cases
increased with the increase of caries depth. This is however
expectable. The increasing trend of diagnostic sensitivity from
enamel to dentin walls are also reported in other studies.19,30
Correct diagnosis of caries depth for enamel caries especially
in the outer surface is low (0e10% for the outer layer and
10.5e44.4% for inner layer). The measured sensitivity for the
enamel decays was shown to be 30.3e66.6% which is more
than the caries depth estimation .This means that enamel
caries are missed in many cases, and in many diagnosed cases
are considered deeper than actual. The exaggerated estimation of enamel caries could be attributed to Mach Band and the
fact that 30e40% of tooth demineralization is necessary for
radiographic detection of caries.1
Caries detection sensitivity for inner half and outer half of
dentin are consecutively 75.7e86.8% and 44.5e68.7%. In cases
of dentin caries like enamel caries, sensitivity was higher
compared to caries depth estimation by observers (56.6%e
20.3%). Results indicate that dentin caries especially those of
inner half are considered shallower than actual. This could be
due to the triangular pattern of caries penetration into dentin;
the base being at the enameledentin connection. Thus in
deeper regions caries depth detection is more difficult due to
the superimposition of bigger proportion of sound dentin.
Furthermore likewise enamel caries, less than 30e40% of
demineralization is only histologically detectable and not
radiographically. Spearman test calculated the intra observers
and inter-observers agreement instead of Kappa test. In
studies of Rockenback and Senel inter-observers agreement
were evaluated good.20,23 In the current study a good agreement was evaluated between all observers (more than 0.71).
Inter-observers agreement for CMOS digital images was reported moderate (0.59e6.7) and for other methods a good
agreement was evaluated (more than 0.7).

5.

Conclusion

Indirect digital images of scanned E- and D-speed films,


standard and filtered direct digital images (CMOS receptor)
have high accuracy. There was no significant difference in the
diagnosis of tooth decay. All 4 study methods exhibit high
validity and reliability for the diagnosis of proximal caries.

Conflicts of interest
All authors have none to declare.

i n d i a n j o u r n a l o f d e n t i s t r y 4 ( 2 0 1 3 ) 8 3 e8 7

references

1. White SC, Pharoah MJ. Oral Radiology Principles and


Interpretation. 6th ed. St Louis: Mosby; 2009: 79e83,
270e381.
2. Yang J, Dutra V. Utility of radiology, laser fluorescence, and
transillumination. Dent Clin North Am. 2005;49:739e752.
3. Pontual AA, Melo de DP, Almeida de SM, Boscolo FN, Haiter
Neto F. Comparison of digital systems and conventional
dental film for the detection of approximal enamel caries.
Dentomaxillofac Radiol. 2010;39:431e436.
4. Pretty CM, Ten Bosch JJ. Dynamics of and diagnostic methods
for detecting small carious lesions. Caries Res.
1996;30:381e388.
5. Seneadza V, Koob A, Kaltschmitt J, Duwenhoegger J,
Eickholz P. Digital enhancement of radiographs for
assessment of interproximal dental caries. Dentomaxillofac
Radiol. 2008;37:142e148.
6. Young DA, Featherstone JD. Digital imaging fiber-optic transillumination, F-speed radiographic film and depth of
approximal lesions. JADA. 2005;136:1682e1687.
7. Castro VM, Katz JO, Hardman PK, Glaros AG, Spencer P.
In vitro comparison of conventional film and direct digital
imaging in the detection of approximal caries. Dentomaxillofac
Radiol. 2007;36:138e142.
8. Geist JR, Brand JW. Sensitometric comparison of speed group
E and F dental radiographic films. Dentomaxillofac Radiol.
2001;30:147e152.
9. Wenzel A. Digital radiography and caries diagnosis (Review).
Dentomaxillofac Radiol. 1998;27:3e11.
10. Langland OE, Langlais RP, Preece JW. Principles of Dental
Imaging. 2nd ed. Philadelphia: Lippincott Williams & Wilkins;
2002: 259e295.
11. Ricketts D, Kidd E, Wilson RF. An evaluation of the
diagnostic yield from bitewing radiographic of small
approximal and occlusal carious lesion in a low prevalence
sample in vitro using different film type and speeds. Br
Dent J. 1997;182:51e58.
12. Eorten H, Akarslan ZZ, Topyz O. The efficiency of three
different films and radiovisiography in the detecting
approximal carious lesion. Quintessence Int. 2005;36:65e70.
13. Syriopoalos K, Sanderink G, Velders XL, et al. Radiographic
detection of approximal caries: a comparison of dental films
and digital imaging systems. Dentomaxillofac Radiol.
2000;29:312e318.
14. Janhom A, Van Ginkel FC, Van Amerongen JP, Van der
Stelt PF. Scanning resolution and the detection of approximal
caries. Dentomaxillofac Radiol. 2001;30:166e171.
15. Langland OE, Langlais RP, Preece JW. Principles of Dental
Imaging. 2nd ed. Batimore: Lippinccot Williams & Wilknis;
2002 [chapter 11]: 259e295.

87

16. Haring JI, Iannucci JM, Jansen L. Dental Radiography: Principles


and Techniques. 2nd ed. Philadelphia: Saunders; 2000 [chapter
6]: 384e394.
17. Moystad A, Svans DB, Van der Stelt PF, et al. Comparison of
standard and task-specific enhancement of Digora storage
phosphor images for approximal caries diagnosis.
Dentomaxillofac Radiol. 2003;32:390e396.
18. Peker I, Alkurt MT, Altunkaynak B. Film tomography compared
with film and digital bitewing radiography for proximal caries
detection. Dentomaxillofac Radiol. 2007;36:495e499.
19. Hintze H, Wenzel A. Influence of the validation method on
diagnostic accuracy for caries. A comparison of six digital and
two conventional radiographic systems. Dentomaxillofac
Radiol. 2002;31:44e49.
, Yuksel SP, O
zen T, Avsever H.
cok O
20. Senel B, Kamburoglu K, U
Diagnostic accuracy of different imaging modalities in
detection of proximal caries. Dentomaxillofac Radiol.
2010;39:501e511.
21. Ludlow JB, Abreu M, Mol A. Performance of a new F-speed film
for caries detection. Dentomaxillofac Radiol. 2001;30:110e113.
22. Hintze H. Diagnostic accuracy of two software modalities for
detection of caries lesions in digital radiographs from four
dental systems. Dentomaxillofac Radiol. 2006;35:78e82.
23. Rockenbach MI, Veeck EB, Costa NP. Detection of proximal
caries in conventional and digital radiographs: an in vitro
study. Stomatologija. 2008;10:115e120.
24. Pereira AC, Eggertsson H, Moustafa A, Zero DT, Eckert GJ,
Mialhe FL. Evaluation of three radiographic methods for
detecting occlusal caries lesions. Braz J Oral Sci. 2009;8:67e70.
25. Ulusu T, Bodur H, Odabas ME. In vitro comparison of digital
and conventional bitewing radiographs for the detection of
approximal caries in primary teeth exposed and viewed by a
new wireless handheld unit. Dentomaxillofac Radiol.
2010;39:91e94.
26. Haak R, Wicht MJ, Noack MJ. Conventional, digital and
contrast-enhanced bitewing radiographs in the decision to
restore approximal caries lesion. Caries Res. 2001;35:193e199.
27. Uprichard KK, Potter BJ, Russell CM, Schafer TE, Adair S,
Weller RN. Comparison of direct digital and conventional
radiography for the detection of proximal surface caries in
the mixed dentition. Pediatr Dent. 2000;22:9e15.
28. Li G, Sanderink GCH, Berkhout WER, Syriopoulos K, Van der
Stelt PF. Detection of proximal caries in vitro using standard
and task-specific enhanced images from a storage phosphor
plate system. Dentomaxillofac Radiol. 2007;41:231e234.
29. Moystad A, Svanaes DB, Risnes S, Larheim TA, Grondahl HG.
Detection of approximal caries with a storage phosphor
system. A comparison of enhanced digital images with dental
X-ray film. Dentomaxillofac Radiol. 1996;25:202e206.
30. Farman AG, Farman TT. A comparison of 18 different X-ray
detectors currently used in dentistry. Oral Surg Oral Med Oral
Pathol Oral Radiol. 2005;99:485e489.

S-ar putea să vă placă și