ARTÍCULOS ORIGINALES

The influence of display modalities on proximal caries detection and treatment decision

 

Véra L.S.A. Barbosa¹, Amanda K.G. Gonzaga¹, Andrea A. Pontual², Patrícia M. Bento¹, Flávia M.M. Ramos-Perez², Pedro T.D. Filgueira¹, Daniela P. Melo¹

¹ Department of Oral Diagnosis, Division of Oral Radiology, State University of Paraíba- UEPB, Campina Grande, Brazil of Pernambuco - UPE, Recife, Brazil.
² Department of Clinical and Preventive Dentistry, Division of Oral Radiology, Federal University of Pernambuco - UFPE, Recife, Brazil.

CORRESPONDENCE Dr. Amanda Katarinny Goes Gonzaga, Rua Manoel Alves de Oliveira, 684 Catole, Campina Grande - PB 58410-575, Brazil; E-mail: amandaggonzaga@gmail.com

---

ABSTRACT

The aim of this study was to investigate the influence of digital radiographic display on caries detection and choice of treatment among undergraduate students. Forty images of extracted human teeth were acquired using a PSP digital system. The proximal surfaces were evaluated for the presence of proximal caries and choice of treatment by 36 undergraduate students, divided into three groups according to the semester they were taking. The images were evaluated in two forms of image display: laptop, and printed on acetate viewed on a lightbox. The accuracy of the different forms of image display on caries detection was evaluated by means of ROC curve analysis and its effect by mixed linear regression. Residue analysis was used to verify the adequacy of the treatment of choice for the chosen diagnosis. There was no significant effect either for the display modalities (p=0.058) or for the different undergraduate student groups (p=0.991). The Az was 0.539 for printed images and 0.516 for laptop. The decisions based on treatment of choice were consistent with the scores achieved for caries detection. Accuracy of caries detection using a laptop was comparable to accuracy using printed images. Treatment decision was not affected by image display modality. The semester of the dentistry course that undergraduate students were taking did not significantly increase the accuracy of their proximal caries detection.

Key words: Dental Caries; Radiography; Dental-Diagnosis; Oral.

RESUMO

Influência dos meios de apresentação radiográfica no diagnóstico e tratamento da cárie

O objetivo deste estudo foi investigar a influencia do meio de apresentacao da imagem radiografica digital no diagnostico da carie e na decisao de tratamento realizado por alunos de graduacao. Foram obtidas 40 imagens digitais de dentes humanos extraidos atraves do sistema digital PSP. As superficies proximais dos dentes foram avaliadas quanto a presenca de carie proximal por 36 estudantes de odontologia, distribuidos em tres grupos de acordo com o nivel de formacao. As avaliacoes foram efetuadas em um laptop e em imagens impressas com o auxilio do negatoscopio. A acuracia dos meios de apresentacao quanto a deteccao de carie incipiente foi avaliada pela media das areas sob as curvas ROC e seu efeito por uma analise de regressao linear mista. Para a tomada de decisao terapeutica foi realizada uma analise de residuos para verificar sua adequacao ao diagnostico. Nao houve efeito significativo nem para a modalidade de visualizacao (p=0.058) e nem para os grupos de alunos (p=0.991). A media das areas sob as curvas Roc para o filme foi de 0.539 e de 0.516 para negatoscopio. A decisao de tratamento foi condizente com o diagnostico efetuado. A acuracia do diagnostico da carie proximal realizado em tela de laptop e comparavel ao realizado em negatoscopio. A tomada de decisao terapeutica nao foi afetada pelo meio de apresentacao da imagem radiografica digital. A progressao do aluno no curso nao melhora a precisao diagnostica da carie proximal.

Palavras chave: Carie Dentaria; Radiografia Dentaria; Diagnostico Bucal.

---

 

INTRODUCTION

Proximal caries still pose a challenge to dental care providers mainly because due to their location, they can only be detected clinically when a great extension of the proximal surface is compromised¹. Bitewing radiography is used for detecting caries, but does not usually detect lesions in early stages, before cavitation. New, more accurate and more reproducible diagnostic methods are therefore needed to supplement early diagnosis and plan appropriate treatment for caries based on their low prevalence and extension and slow progression². The use of dental radiographic films has largely been replaced by intraoral digital radiographic systems, most of which have been tested for efficiency in detecting caries^3-7. Because digital images enable the acquired image to be viewed on a computer monitor, the evaluation of the different forms of image displays is of interest for dental health care providers and researchers. Haak et al.⁸
evaluated the influence of digital image size on different types of monitors and found that the type of monitor has no effect on caries detection. Another form of image display still used by some professionals is a printed image, usually on acetate, which is similar to a conventional image8. Hellen-Halme et al.⁹ suggest that in order to view proximal caries in digital images, ambient light conditions should be dimmed and brightness and contrast of the monitor adjusted to provide excellent image quality. Carmona et al.¹⁰ claim that observer experience improves radiographic caries detection, reducing false negative results and increasing accuracy. Based on the above, the aim of this study is to evaluate the influence of different image displays and the level of experience of undergraduate students on the diagnosis of proximal caries and treatment decisions by observing dental students from a Brazilian institution.

MATERIALS AND METHODS

This study was approved by the Ethics Committee of the Department of Dentistry at the State University of Paraiba (Protocol number CAAE 0384.0.133.000- 11). Fifty extracted human teeth (10 canines, 20 premolars and 20 molars) were mounted on 10 silicone blocks, each of which held five teeth, simulating a normal condition. The canine was used to ensure proximal contact with the first premolar. The premolars and molars were either intact or had a small area of demineralization on their proximal surfaces. Digital images were acquired using a GE 1000 (General Electric Company, Milwaukee, WI, USA) unit operating at 65 kVp and 10 mA. An acrylic plate 1.2 cm thick was placed adjacent to the models as a material equivalent to soft tissue. To ensure reproducible imaging geometry, the silicone blocks were stabilized on a customized acrylic device to provide a distance of 34 cm between the target and the image receptor, a centrally oriented X-ray beam and a distance of 2 cm between the teeth and the receptor.
The image receptor used in this study was size 2 intraoral PSP digital imaging (DenOptix, Gendex Dental Systems, Milan, Italy). Before exposure, each plate of the digital system was exposed to a lightbox for 130 s, as recommended by the manufacturer. The exposed phosphor plates were scanned using 300 dpi resolution and the files were exported and saved in tagged image file format (TIFF). The acquired TIFF images were mounted in a PDF and printed using AGFA acetate film (Agfa Healthcare, Gevaert Group, Belgium) on a laser printer Drystar 5300 (Agfa Healthcare, Gevaert Group, Belgium). Each film contained eight images, and there were a total of five films.

Gold Standard
Histological sections (700 μm) served as a validating criterion for the presence and depth of the caries lesions. Before selection, the teeth were individually embedded in acrylic (Vipi, Sao Paulo, Brazil) and then sectioned in mesiodistal direction using a 200 mm diamond band. The sections were cleaned and glued to microscope slides using transparent varnish. Independent histological validation was performed by two previously trained observers under incident light (12.5 - 20X magnification) using a binocular microscope. If the observers' ratings varied, they were asked to perform a joint assessment to establish agreement. Caries were defined as present when an opaquewhite demineralization or brown discolored area was observed on the surface. For the histological surface, the following scale was applied: 0 = no enamel demineralization or narrow surface zone of opacity; 1 = demineralization limited to the enamel; 2 = demineralization involving the dentine.

Viewing Sessions
Prior to the examination sessions, the observers received explanations and practical instructions, and underwent calibration tests so that they would be familiar with the specific characteristics of the digital images used in the study. Viewers need to be trained with regard to the specific characteristics of images in each digital system so that diagnosis would not be compromised by the difference between systems. The professors were not required to undergo calibration because they are already familiar with the digital images used in this institution. The researchers who were not familiar with them underwent calibration but did not participate in the student training. During calibration, the evaluation method was explained, and training and knowledge were verified. The researcher responsible for training the observers remained in the same room to answer any questions that might arise during the evaluation sessions. The digital images were displayed on a 17" color laptop monitor, size 1:1, and the printed images were displayed on a lightbox placed in a quiet room with dimmed lighting. The digital images could not be enhanced. All images were evaluated in a quiet and darkened room. Each observer evaluated individually forty teeth, resulting in the evaluation of eighty proximal surfaces. The number of images evaluated at each session was determined by the observer and could not exceed 20 at a time.
Thirty-six independent observers, all undergraduate dentistry students at the State University of Paraiba (14 from the 7th semester, 12 from the 8th semester, and 10 from the 9th semester) were selected to evaluate the images. They were chosen according to their grades and success in radiology classes, and were considered able to diagnose dental caries. They were divided into groups based on the semester they were taking and coded as: Group A - 7th semester, Group B - 8th semester, and Group C - 9th semester. The presence of proximal caries lesions was scored using a 5-point confidence scale: 1 = definitely not present, 2 = probably not present, 3 = unsure, 4 = probably present, and 5 = definitely present. After scoring the images individually for the presence of caries lesion, the observer scored the same image to indicate adequate clinical follow-up as: N - no treatment needed, P - Preservation, R - Restorative treatment.

Data Analysis
To measure the accuracy of the images displayed in different modalities and evaluated by different groups, the area under the Receiver Operating Characteristic (ROC) curves (Az) were calculated. Az was calculated for each observer in each modality evaluated. The possible effect of the observer group and image display modality on the ROC curves was evaluated using a mixed linear regression model, taking into consideration the possible structure of score correlation because each observer evaluated the dental surface in two forms of image visualization: printed on acetate viewed on a lightbox, and on a laptop monitor. The possible association between caries detection and treatment decision was verified using the chi-square test. Adjusted residue analysis was used to identify the most significant sources of association presented in each contingency table for each group. All statistical tests adopted a 0.05 level of significance.

RESULTS

Of the 80 microscopically evaluated surfaces, 29 (36.25%) presented proximal caries lesions, of which 26 were restricted to the enamel, and only 3 reached the outermost dentin. Table 1 shows the mean values for the area under the ROC curve, standard deviations, and confidence intervals (CI) for each group of observers for each image display modality. There was no statistical difference between display modalities (p = 0.058) or among the three observers groups (A, B, C) (p = 0.991). Tables 2a and 2b show the adequacy of treatment choice to the caries detection scores chosen by group A for printed and laptop monitor image modality, respectively. The restorative treatment

 

decision for surfaces scored for caries definitely present was practically unanimous (98.5% and 98.7%), whereas surfaces considered sound did not receive indication for treatment (92.8% and 98%) for either image display modality.

Table 1: Means and Standard deviations of the areas under the ROC curves for the undergraduate student groups and image modalities.

Table 2a: Caries detection and treatment decision scores for the 14 observers in Group A for display modalities.

Table 2b: Adjusted residues for Group A on display modalities.

Tables 3a and 3b represent the correspondence between caries detection and treatment of observer group B on both image display modalities. In this group there is greater consistency between caries detection and restorative treatment decision when evaluating image modality (100%). None of the faces scored as sound or probably sound were scored for restorative treatment. Tables 4a and 4b show the results for therapy chosen based on the diagnosis made by group C. Surfaces that the observers considered questionable were indicated for follow-up when evaluated in printed image modality by all observers, whereas for the laptop monitor modality, the result was 96.4%.

Table 3a: Caries detection and treatment decision scores for the 14 observers in Group B for display modalities.

Table 3b: Adjusted residues for Group B on display modalities.

Table 4a: Caries detection and treatment decision scores for the 14 observers in Group C for display modalities.

Table 4b: Adjusted residues for Group C on display modalities.

DISCUSSION

This study evaluated the influence of two different image display modalities on caries detection and treatment choice. Two methods for displaying the radiographic digital image were compared: printed on acetate viewed on a lightbox, and displayed on a laptop monitor. This study also aimed to evaluate whether students' undergraduate dental experience in radiology classes and clinical practice improved their capacity to detect incipient proximal caries lesions. Most of the studies which have evaluated the influence of image quality on caries detection used a relatively small number of observers, typically oral radiologists or dental professionals with significant experience in caries detection^4,11,12. In Rockemback et al. ¹³, one observer conducted the evaluation, while in other studies the number of observers varied from three^11,14,15 to 143 or 2016. Few studies used dental students as observers^16-18. In our study, the evaluations were performed by 36 dental students in order to investigate the influence of academic training and level of experience on radiographic proximal caries detection. The teeth sample used in this study consisted of premolars and molars with carious lesion in the initial stages, which previous studies found to be more difficult to detect^{3,9,11,12,19,20}. According to Shintaku, Scarbeczm and Venturim, the Az values are significantly higher when results are validated based only on radiographic and clinical evaluation^21,22.
To assess how the scores reflected on the diagnostic decision, Az values were calculated for each observer for each modality of image display. As in our study, Ferreira et al. evaluated enamel demineralization using radiographic film and digital and digitized images²³. They evaluated the diagnostic performance of each observer and measured the efficiency of the methods by using the areas under the ROC curves. Otis & Sherman²² evaluated caries diagnosis using printed photographic paper interproximal images and in the statistical analysis, calculated Az for each observer and image format. Other similar studies evaluated observer performance by using ROC curve analysis^{4-6,11,16,22,24}. Ludlow & Abreu²⁵ found that the accuracy values for caries lesions evaluated on laptop monitors were comparable to the accuracy values for printed images evaluated using the lightbox, in agreement with our study. All viewing sessions in our study were done in a quiet room with dimmed light for both image display methods; a setup which substantially increases diagnostic precision^9,12,26. The Az projects total diagnostic accuracy and consequently represents the performance of the observer. In this type of analysis, perfect performance is represented by an Az value close to 1; nevertheless, values from 0.75 to 0.80 are considered acceptable for the detection of proximal caries involving dentin tissue. However, incipient caries are more difficult to diagnose because enamel decalcification needs to be higher than 30% for them to be detected in radiographic images, leading to lower Az levels than when evaluating dentin lesions¹⁰. In this study, the Az values ranged from 0.515 to 0.542, revealing lower diagnostic accuracy than is usually found in studies of this nature. Similar Az values were found in a study by Diniz et al.¹⁶, in which the Az value for Brazilian undergraduate dental students was 0.53 for enamel caries detection. Hellen- Halme, Nilsson and Petersson4 reported similar Az values, ranging from 0.542 to 0.557, even though the observers were one oral radiologist and six dental professionals, including dentists with more professional experience than the observers in our study, who were undergraduate dentistry students.
Adjusted residual analyses based on caries detection and therapeutic treatment decision for both forms of image display reflected alignment of academic teaching at the university with the new caries treatment philosophy. The strong tendency to prefer preservation over restorative treatment was clearly evident. The score showed that restoration was only selected for surfaces scored as caries definitely present. The theory of minimum intervention is based on preventive dentistry, which postpones filling or its substitution until strong evidence exists of caries lesion²⁷ and recommends maximum preservation of sound tooth structure, executing conservative preparation and providing higher resistance to the remaining dental structure. New developments in long-lasting restorative adhesive materials enable the use of conservative preparation limited to the size of the lesion and do not require additional time wearing of mechanical retention²⁸.
Rocha et al. found that, when in doubt, the undergraduate student prefers to classify the surfaces as intact during radiographic evaluation, increasing the number of false negatives, which immediately reflects on the therapeutic choice in the form of nontreatment or preservation¹⁷. On the other hand, the results found by Bervian et al.¹⁸ revealed that a significant proportion of students chose restorative treatment for caries lesions restricted to the enamel surface, and this proportion was more evident among students attending private schools. As previously seen in other professional courses, the quality of dentistry education is related to the university and the course having an adequate pedagogical model . Despite the occurrence of major changes in higher education, the educational model for dentistry in Brazil still focuses mainly on technical training. Our study found that when the diagnosis was either caries definitely present or definitely sound, the treatment decision was consistent with the diagnosis, which implies greater confidence of the observer during evaluation of the images. Definitely sound surfaces were not treated, while restorative treatment was indicated when caries were definitely present. The uncertain options, considered probably present or probably sound, or without any condition to diagnose -score 3-, led to follow-up, which is a safe, conservative therapy. These students should be trained to detect incipient proximal caries more effectively because the definitive diagnosis should not be jeopardized in favor of follow-up, as a neglected caries lesion can evolve. In conclusion, this study demonstrated that the different forms of digital image displays resulted in similar performance by observers and did not affect the precision of proximal caries detection or treatment decision. Student level in a dentistry course does not significantly increase precision with regard to proximal caries detection.

REFERENCES

1. Hala LA, Mello JB de, Carvalho PL de. Evaluation of the effectiveness of clinical and radiographic analysis for the diagnosis of proximal caries for different clinical experience levels: comparing lesion depth through histological analysis. Braz J Oral Sci 2006;5:1012-1017.         [ Links ]

2. Tsuchida R, Araki K, Okano T. Evaluation of a limited cone-beam volumetric imaging system: comparison with film radiography in detecting incipient proximal caries. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2007; 104:412-416.         [ Links ]

3. Shi XQ, Li G. Detection accuracy of approximal caries by black-and-white and color-code digital radiographs. Oral Surg Oral Med Oral Pathol Oral Radiol Oral Endod 2009; 107:433-436.         [ Links ]

4. Hellen-Halme K, Nilsson M, Petersson A. Effect of monitors on approximal caries detection in digital radiographsstandard versus precalibrated DICOM part 14 displays: an in vitro study. Oral Surg Oral Med Oral Pathol oral radiol Endod 2009;107:716-720.         [ Links ]

5. Araki k, Matsuda Y, Seki K, Okano T. Effect of computer assistance on observer performance of approximal caries diagnosis using intraoral digital radiography. Clin Oral Investig 2010;14:319-325.         [ Links ]

6. Li G, Qu XM, Chen Y, Zhang J, Zhang ZY, Ma XC. Diagnostic accuracy of proximal caries by digital radiographs: an in vivo and in vitro comparative study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2010;109:463-467.         [ Links ]

7. Kamburoglu K, Senel B, Yuksel SB, Ozen T. A comparison of the diagnostic accuracy of in vivo and in vitro photostimulable phosphor digital images in the detection of occlusal caries lesions. Dentomaxillofac Radiol 2010; 39:17-22.         [ Links ]

8. Haak R, Wicht MJ, Nowak G, Hellmich M. Influence of displayed image size on radiographic detection of approximal caries. Dentomaxillofac Radiol 2003;32: 242-246.         [ Links ]

9. Hellen-Halme K, Petersson A, Warfvinge G, Nilsson M. Effect of ambient light and monitor brightness and contrast settings on the detection of approximal caries in digital radiographs: an in vitro study. Dentomaxillofac Radiol 2008;37:380-384.         [ Links ]

10. Carmona GP, Devito KL, Pontual MLA, Haiter-Neto F. Influencia da experiencia profissional no diagnostico radiografico de caries. Cienc Odontol Bras 2006;9:87-92.         [ Links ]

11. 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;32:504-509.         [ Links ]

12. Hellen-Halme K, Lith A. Effect of ambient light level at the monitor surface on digital radiographic evaluation of appoximal carious lesions: an in vitro study. Dentomaxillofac Radiol 2012;41:192-196.         [ Links ]

13. Rockenbach MI, Veeck EB, da Costa NP. Detection of proximal caries in conventional and digital radiographs: an in vitro study. Stomatologija 2008;10:115-120.         [ Links ]

14. Erten H, Akarslan ZZ, Topuz O. The efficiency of three films and radiovisiography in detecting approximal carious lesions. Quintessence Int 2005;36: 65-70.         [ Links ]

15. Hintze H. Diagnostic accuracy of two software modalities for detection of caries lesions in digital radiographs from four dental systems. Dentomaxillofac Radiol 2006;35:78-82.         [ Links ]

16. Diniz MB, Rodrigues JA, Neuhaus KW, Cordeiro RC, Lussi A. Influence of examiner's clinical experience on the reproducibility and accuracy of radiographic examination in detecting occlusal caries. Clin Oral Investig 2010;14: 515-523.         [ Links ]

17. Rocha AS, Almeida SM, Boscolo FN, Haiter Neto F. Interexaminer agreement in caries radiographic diagnosis by conventional and digital radiographs. J Appl Oral Sci 2005;13:329-333.         [ Links ]

18. Bervian J, Tovo MF, Feldens CA, Brusco LC, Rosa FM. Evaluation of final-year dental students concerning therapeutic decision making for proximal caries. Braz Oral Res 2009;23:54-60.         [ Links ]

19. Schulze D, Heiland M, Thurmann H, Adam G. Radiation exposure during midfacial imaging using 4- and 16-slice computed tomography, cone beam computed tomography systems and conventional radiography. Dentomaxillofac Radiol 2004;33:83-86.         [ Links ]

20. Pontual AA, de Melo DP, de Almeida SM, Boscolo FN, Haiter Neto F. Comparison of digital systems and conventional dental film for the detection of approximal enamel caries. Dentomaxillofac Radiol 2010; 39:431-436.         [ Links ]

21. Shintaku WH, Scarbecz M, Venturin JS. Evaluation of interproximal caries using the IPad 2 and a liquid crystal display monitor. Oral Surg Oral Med Oral Pathol Oral Radiol 2012;113:e40-44.         [ Links ]

22. Otis LL, Sherman RG. Assessing the accuracy of caries diagnosis via radiograph: Film versus print. J Am Dent Assoc 2005;136:323-330.         [ Links ]

23. Ferreira RI, Haiter-Neto F, Tabchoury CP, de Paiva GA, Boscolo FN. Assessment of enamel demineralization using conventional, digital, and digitized radiography. Braz Oral Res 2006;20:114-119.         [ Links ]

24. Li G, Sanderink GC, Berkhout WE, Syriopoulos K, van der Stelt PF. Detection of proximal caries in vitro using standard an task-specific enhanced images from a storage phosphor plate system. Caries Res 2007;41:231-234.         [ Links ]

25. Ludlow JB, Abreu M Jr. Performance of film, desktop monitor and laptop displays in caries detection. Dentomaxillofac Radiol 1999;28:26-30.         [ Links ]

26. Kutcher MJ, Kalathingal S, Ludlow JB, Abreu M Jr, Platin E. The effect of lighting conditions on caries interpretation with a laptop computer in a clinical setting. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2006; 102:537-543.         [ Links ]

27. Anusavice KJ. Does ART have a place in preservative dentistry? Community Dent Oral Epidemiol 1999;27:442-448.         [ Links ]

28. Tyas MJ, Anusavice KJ, Frencken JE, Mount GJ. Minimal intervention dentistry-a review. FDI Commission Project 1-97. Int Dent J. 2000;50:1-12.         [ Links ]