Original Article

In vitro comparison of two photostimulable phosphor plate systems for early detection of occlusal dentin caries with and without a sharpening filter

Introduction

Dental carries, as a progressive bacterial disease, is one of the most common diseases, affecting 95% of the population; it is believed to be the most important reason for tooth loss. ¹ Unfortunately, no accurate and sensitive tool is available to help diagnose dental caries in its initial stages. ² Although different techniques, such as analog and digital radiography, transillumination, fluorescence, and tomography, are useful for the diagnosis of incipient caries, radiography is still the most commonly used technique for the diagnosis of caries. ³ Based on previous studies, 25‒42% of carious lesions remain undetected during clinical examinations without the use of radiographic techniques. ⁴ There is controversy over the diagnostic power of radiography for carious lesions. ⁵ Some researchers believe that the diagnostic accuracy of E- and F-speed films is similar to that of digital radiography for proximal caries. ⁶ Pereira et al ⁷ reported that considering the advantages of digital radiography, it appears it is rational to replace digital imaging systems for conventional radiographic systems, even with a diagnostic accuracy similar to that of conventional radiography. Many studies have evaluated the diagnostic power of photostimulable phosphor plate (PSP), CMOS, and CCD digital systems for detecting proximal caries. ^8,9 Contrary to proximal caries, the diagnostic accuracy for occlusal caries is a matter of controversy, despite the fact that determination of the role of caries progression in the enamel and dentin depth is very important for preparing a correct treatment plan. ⁹ Therefore, the evaluation of these diagnostic techniques can help dentists select the best diagnostic system for the diagnosis of occlusal caries. ⁷ Studies comparing the image quality of phosphor plates with conventional films and the CCD systems have reported a comparable or similar image quality for phosphor plates and a wide dynamic range and higher contrast and resolution with lower exposure doses for PSP. ¹⁰

Controlled clinical and laboratory studies are necessary to determine whether these new digital systems with image enhancement capabilities improve diagnosis, treatment, and prognosis. ¹¹

Considering the paucity of studies on the subject, the present in vitro study was undertaken to evaluate and compare the accuracy of Digora and Acteon PSP digital systems in the diagnosis of occlusal caries with and without a sharpening filter.

Methods

In this experimental study, 69 extracted human molar and premolar teeth, with no visible occlusal cavities, restorations, or Cl V cavities, were evaluated. The teeth were stored in 10% formalin, cleaned with water spray, and dried with an air syringe before being used. In the next stage, the teeth were mounted in blocks of stone and sawdust in rows consisting of three samples. Then the teeth were numbered and underwent a radiographic procedure under standard conditions with the use of Minray unit (Helsinki, Finland) with two different intraoral PSP systems of Digora Optime (Sordex, Helsinki, Finland) and Acteon (Soppro, La Ciotat Cedex France) at kVp=70, mA=8, and an exposure time of 0.2 seconds, already determined in a pilot study. The tooth blocks were placed in a film holder so that the radiographic procedures were carried out under similar geometric conditions. The film-to-tube distance was set at 41 cm, and a piece of Plexiglass was placed between the tooth blocks and the tube to simulate soft tissues. The images were captured with Scanora software and saved with numbers; sharpening filter 1 was then applied to all the images and saved with other numbers (). All the images were displayed randomly on a Samsung monitor (Sync Master 740 N) and evaluated by two experienced observers to detect caries twice at an interval of two weeks. The observers were permitted to manipulate images to change the image density and contrast.

Close

Figure 1. Examples of original (A) and sharped (B) images captured with the Acteon system.

The findings reported by the observers were recorded in Tables prepared for this purpose and consisted of the following:

R0 = no caries

R1 = dentin caries

Caries odds:

1 = definitive absence of caries

2 = caries possibly absent

3 = cannot be determined

4 = caries possibly present

5 = caries definitively present

After radiographic imaging, the tooth samples were retrieved from the blocks and re-mounted in single acrylic resin blocks and numbered. The teeth were sectioned in the buccolingual direction along the vertical axis of each tooth with a Mecatome machine (T201A) (PRESI Co., France) at low speed using a diamond saw (the cutting edge of the blade was made of diamond with a thickness of 0.5 mm). Two or three sections were prepared from each tooth, measuring 1000 µm in thickness.

Subsequently, the tooth sections were viewed under a stereomicroscope using magnification by a pathologist (), and the sound and carious occlusal enamel and dentin were recorded in tables prepared to this end. Finally, the observers’ diagnoses were compared with the histopathological diagnosis as the gold standard.

Close

Figure 2. A sample of cross-sections under a stereomicroscope.

Analysis of data

Data were analyzed with SPSS 23. Kappa analysis was used to evaluate intra- and inter-observer agreements. A kappa coefficient of ≥0.8 was considered excellent agreement, with 0.6‒0.79 as good, 0.40‒0.59 as moderate, 0.20‒0.39 as poor, and <0.2 as very poor agreement. To evaluate sensitivity and specificity, a 5-scale table was convened to a 2-scale table so that the values 1 and 2 (caries definitively present and caries possibly present) were considered as the presence of caries, and three other scores were considered as the absence of caries. Z test was used to compare sensitivity and specificity. To evaluate the accuracy, the surface area under the ROC curve (AUC) was used. AUC>0.9 was considered excellent accuracy, with 0.8‒0.9 being considered good, 0.7‒0.8 as moderate, and 0.6‒0.7 as poor accuracy ().

Close

Figure 3. The surface area under the ROC curve for dentin caries separately for each system.

Results

Histopathological evaluation of 69 teeth in the present study showed that 14 teeth (20.3%) were sound, with 25 teeth (46.2%) and 30 teeth (43.5%) exhibiting enamel and dentin caries, respectively. Intra- and inter-observer agreement results are presented in Table 1. Intra-observer agreement for the first observer was between 0.541 and 1.00, with 0.641 and 1.00 in the second observer. The lowest agreement was related to the Acteon system in the sound teeth. The inter-observer agreement was between 0.451 and 1.00. The lowest agreement was related to the first observation of sound teeth in the Acteon system (Table 1).

The sensitivity of both systems was 28.3 to 34.1, with a specificity of 92.9‒94.8 (Table 2).

The accuracy [(true positive + true negative)/total observations] of both systems in two filtering modes was measured and compared with the AUC and the related confidence intervals Table 2); the difference in AUC was not significant [0.645 (CI: 0.578‒0.713) vs. 0.659 (CI: 0.592‒0.762), 0.649 (CI: 0.581‒0.716) and 0.674 (CI: 0.608‒0.704)]. According to Table 3, this difference was not significant concerning the observer.

Discussion

Several studies have evaluated the effects of manipulating digital images and different filters on the diagnostic accuracy of images, with different results. Based on some studies, the manipulation of images and the use of different filters such as ‘sharp’ does not affect on increasing the caries diagnostic power of digital radiographic techniques. ¹² Studies by Belem et al ¹² and Kositbowornchai et al ¹³ reported such a result. In the present study, the observers were allowed to manipulate images to change contrast, density, and magnification. Besides, the effect of sharpening filter 1 on the diagnosis of occlusal caries was evaluated.

Intra- and inter-observer agreement in both systems for original and sharpened images was good to excellent, except for star-marked sound surfaces in the Acteon system (Table 1). The agreement rates in the study carried out by Rocha et al ¹⁴ were 0.51 and 0.31 in dental students and radiologists. The kappa coefficients in the study carried out by Shams et al ¹⁵ in third-year students, last-year students, postgraduate radiology students, and general dental practitioners were 0.002, 0.073, 0.271, and 0.03, respectively, which are lower in both studies compared to the present study.

The differences in internal and external agreement rates between different studies might be attributed to the following: 1) the experience of the observers with digital systems; 2) the diagnostic capabilities of the observers concerning caries; 3) the time interval between the observations; 4) the number of observers.

Based on Table 2, the sensitivity of both systems was 28.3 to 34.1, with no significant difference between the two systems (P > 0.3); the accuracy of early diagnosis of occlusal caries was poor in both systems, and no significant difference was observed between the two systems at 95% confidence interval. These results were similar to some previous studies. ^7,16-18

Although sensitivity and the AUC were slightly higher in the ‘unsharp’ mode compared to the ‘sharp’ mode, there was no significant difference between them. Filters that sharpen an image enhance boundaries with high-frequency noise removal; therefore, sharpening filters remove grey scales that might have a diagnostic value in the detection of incipient caries.

In Shokri et al ¹⁹ study, the sensitivity and accuracy of filtered images were significantly higher than original images; this difference was higher in superficial images. Caries in this study was artificially created chemically and had more regular outlines; this affects caries detection on radiographs.

Experience with digital systems strongly affects the results of such studies. ²⁰ Shams et al ¹⁵ evaluated the effect of experience and education on the diagnosis of proximal caries in 2015 and concluded that although experience and knowledge are effective in improving the accuracy of detecting caries on digital images, it does not increase the diagnostic accuracy to the optimal level. Mileman et al ²¹ and Rocha et al ¹⁴ evaluated the effect of experience on the diagnostic accuracy of occlusal and proximal caries and reported that inexperienced students exhibited the highest false positive rate, and radiologists exhibited the highest false negative rate in their reports. In the present study, two radiologists evaluated the images, and consistent with the studies above, there were more false-negative reports than false-positive reports.

In studies by Wenzel et al, ²² Hintze et al, ²³ Rocha et al, ¹⁴ Hintze, ²⁴ Yalçinkaya et al, ²⁰ and Tarım Ertas et al ²⁵ to compare different digital systems and conventional films, it was concluded that there was no significant difference between the accuracy of different radiographic systems, consistent with the results of the present study. Tyndall et al ²⁶ evaluated the effect of manipulating the contrast and density of digital images on their diagnostic efficacy and concluded that the manipulated images exhibited significantly lower accuracy than conventional radiographs and un-manipulated digital images. They used a CCD digital system (Sidexis) for imaging procedures and reported that Sidexis systems use a special processing filter, which improves the images before displaying them on the monitor, and this processing filter might interfere with changes in contrast and illumination, decreasing the diagnostic accuracy in manipulated images. Also, inadequate training and incorrect use of the software program used to manipulate digital images are considered other reasons for a decrease in the accuracy of the above-mentioned manipulated images.

The mean surface area under the ROC curve for the occlusal surface in the study carried out by Wenzel et al ²² was 0.873 for different digital systems, which is higher than the present study. In the study by Wenzel et al, ²² teeth with occlusal cavities were not excluded from the study.

The mean surface area under the ROC curve in the study carried out by Hintze ²⁴ in different digital systems was approximately 0.7, which is higher than that in the present study. In this study, enamel and dentin caries on the proximal surface and dentin caries on occlusal surfaces were evaluated, and both surfaces were evaluated simultaneously to calculate diagnostic accuracy; however, in the present study, proximal caries was not evaluated.

Conclusion

The present in vitro study, with small sample size, showed that the diagnostic accuracy of two PSP sensors of Acteon and Digora systems was the same for the diagnosis of occlusal caries. Although both systems exhibited low sensitivity for the detection of early dentin caries, their high specificity can prevent unnecessary procedures in the clinic. There was no significant difference between images enhanced with enhancement filter 1 and original images.

Authors’ Contributions

Concept or design of the study: RB. Data acquisition, analysis, or interpretation: MM, MM. Writing the manuscript: RB. Critical revision of the article: RB, MM. Final approval of the manuscript: RB, MM, MM. Overall responsibility: RB.

Acknowledgements

The authors would like to thank all the colleagues who sincerely offered their assistance for carrying out this study.

Funding

None.

Competing Interests

The authors declare no competing interests with regards to the authorship and publication of this article.

Ethical Approval

Not applicable.

References

1. Zangooei Booshehry M, Davari A, Ezoddini Ardakani F, Rashidi Nejad MR. Efficacy of application of pseudocolor filters in the detection of interproximal caries. J Dent Res Dent Clin Dent Prospects 2010; 4(3):79-82. doi: 10.5681/joddd.2010.020 [Crossref]
2. Abesi F, Mirshekar A, Moudi E, Seyedmajidi M, Haghanifar S, Haghighat N. Diagnostic accuracy of digital and conventional radiography in the detection of non-cavitated approximal dental caries. Iran J Radiol 2012; 9(1):17-21. doi: 10.5812/iranjradiol.6747 [Crossref]
3. Marami A, Noormohammadi R, Javaheri M. In vitro comparison of the accuracy of Primax, Insight Dental X-Ray films and CMOS-APS digital imaging in detection of interproximal caries. Journal of Shahid Sadoughi University of Medical Sciences 2009; 16(4):33-40.
4. Mehralizade S, Nemati Anaraki S, Sakhdari S, Miraba S, Amiri Siavashani M, Bayat S. Comparing the diagnostic accuracy of two different resolutions of radiographs captured with PSP digital intraoral receptors in detection of secondary caries (in vitro). J Res Dent Maxillofoc Sci 2016; 1(1):40-5. doi: 10.29252/jrdms.1.1.40 [Crossref]
5. Talaeipour A, Nemati Anaraki S, Rezaei Afshar S, Bayat S. Comparing between the power of CCD and PSP digital intraoral receptors in detection of secondary proximal caries in permanent premolar teeth (In Vitro). J Res Dent Sci 2016; 12(4):176-81.
6. Alkurt MT, Peker I, Bala O, Altunkaynak B. In vitro comparison of four different dental X-ray films and direct digital radiography for proximal caries detection. Oper Dent 2007; 32(5):504-9. doi: 10.2341/06-148 [Crossref]
7. 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(2):67-70. doi: 10.20396/bjos.v8i2.8642204 [Crossref]
8. Haiter-Neto F, dos Anjos Pontual A, Frydenberg M, Wenzel A. A comparison of older and newer versions of intraoral digital radiography systems: diagnosing noncavitated proximal carious lesions. J Am Dent Assoc 2007; 138(10):1353-9. doi: 10.14219/jada.archive.2007.0052 [Crossref]
9. Dias da Silva PR, Martins Marques M, Steagall W Jr, Medeiros Mendes F, Lascala CA. Accuracy of direct digital radiography for detecting occlusal caries in primary teeth compared with conventional radiography and visual inspection: an in vitro study. Dentomaxillofac Radiol 2010; 39(6):362-7. doi: 10.1259/dmfr/22865872 [Crossref]
10. Dalili Z, Taramsari M, Mohtavipour ST, Khademi J, Soleymani F, Salamat F. The Effect of delayed scanning of PSP sensors in different storage conditions on the diagnostic quality of apical portion. J Mashhad Dent Sch 2014;38(2):107-18. [Persian].
11. Mirshekar A, Varshosaz M, Fallah Z, Amoozadeh Omran M, Shahnavazi M. Comparison of three digital image receptors and their soft wares in diagnosis of interproximal dental caries (an in vitro study). Journal of Mazandaran University of Medical Sciences 2012;22(93):104-10. [Persian].
12. Belém MD, Ambrosano GM, Tabchoury CP, Ferreira-Santos RI, Haiter-Neto F. Performance of digital radiography with enhancement filters for the diagnosis of proximal caries. Braz Oral Res 2013; 27(3):245-51. doi: 10.1590/s1806-83242013000300004 [Crossref]
13. Kositbowornchai S, Basiw M, Promwang Y, Moragorn H, Sooksuntisakoonchai N. Accuracy of diagnosing occlusal caries using enhanced digital images. Dentomaxillofac Radiol 2004; 33(4):236-40. doi: 10.1259/dmfr/94305126 [Crossref]
14. Rocha AS, Almeida SM, Bóscolo FN, Haiter Neto F. Interexaminer agreement in caries radiographic diagnosis by conventional and digital radiographs. J Appl Oral Sci 2005; 13(4):329-33. doi: 10.1590/s1678-77572005000400003 [Crossref]
15. Shams N, Panahandeh N, Aghababa H, Shams B, Hemati E. Effects of education and experience on detection of proximal caries on digital radiographs. Oral Radiol 2016; 32(3):154-9. doi: 10.1007/s11282-015-0227-3 [Crossref]
16. Mestriner SF, Vinha D, Mestriner Junior W. Comparison of different methods for the occlusal dentine caries diagnosis. J Appl Oral Sci 2005; 13(1):28-34. doi: 10.1590/s1678-77572005000100007 [Crossref]
17. Alomari QD, Qudeimat MA, Ghayyath AA. Imaging of occlusal dentine caries: a comparison among conventional radiographs, digital radiographs, and cone-beam computed tomography images. Oral Radiol 2015; 31(2):73-80. doi: 10.1007/s11282-014-0181-5 [Crossref]
18. Krzyżostaniak J, Kulczyk T, Czarnecka B, Surdacka A. A comparative study of the diagnostic accuracy of cone beam computed tomography and intraoral radiographic modalities for the detection of noncavitated caries. Clin Oral Investig 2015; 19(3):667-72. doi: 10.1007/s00784-014-1282-6 [Crossref]
19. Shokri A, Kasraei S, Lari S, Mahmoodzadeh M, Khaleghi A, Musavi S. Efficacy of denoising and enhancement filters for detection of approximal and occlusal caries on digital intraoral radiographs. J Conserv Dent 2018; 21(2):162-8. doi: 10.4103/jcd.jcd_213_17 [Crossref]
20. Yalçinkaya SE, Künzel A, Becker J. In vitro comparison of intraoral films and four image plate systems in radiographic caries diagnosis. Clin Exp Health Sci 2012; 2(4):175-80. doi: 10.5455/musbed.20121231011629 [Crossref]
21. Mileman PA, van den Hout WB. Comparing the accuracy of Dutch dentists and dental students in the radiographic diagnosis of dentinal caries. Dentomaxillofac Radiol 2002; 31(1):7-14. doi: 10.1038/sj/dmfr/4600652 [Crossref]
22. Wenzel A, Borg E, Hintze H, Gröndahl HG. Accuracy of caries diagnosis in digital images from charge-coupled device and storage phosphor systems: an in vitro study. Dentomaxillofac Radiol 1995; 24(4):250-4. doi: 10.1259/dmfr.24.4.9161170 [Crossref]
23. Hintze H, Wenzel A, Frydenberg M. Accuracy of caries detection with four storage phosphor systems and E-speed radiographs. Dentomaxillofac Radiol 2002; 31(3):170-5. doi: 10.1038/sj/dmfr/4600686 [Crossref]
24. Hintze H. Diagnostic accuracy of two software modalities for detection of caries lesions in digital radiographs from four dental systems. Dentomaxillofac Radiol 2006; 35(2):78-82. doi: 10.1259/dmfr/50356588 [Crossref]
25. Tarım Ertas E, Küçükyılmaz E, Ertaş H, Savaş S, Yırcalı Atıcı M. A comparative study of different radiographic methods for detecting occlusal caries lesions. Caries Res 2014; 48(6):566-74. doi: 10.1159/000357596 [Crossref]
26. Tyndall DA, Ludlow JB, Platin E, Nair M. A comparison of Kodak Ektaspeed Plus film and the Siemens Sidexis digital imaging system for caries detection using receiver operating characteristic analysis. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998; 85(1):113-8. doi: 10.1016/s1079-2104(98)90408-5 [Crossref]