Comparison of the Ability of Two Brands of CBCT with That of SEM to Detect the Marginal Leakage of Class V Composite Resin Restorations

Objectives This study was carried out to compare the ability of two common brands of cone-beam computed tomography (CBCT), including New Tom and Planmeca, to detect the marginal leakage of class V composite resins. The ability of each of the two brands of CBCT to detect the marginal leakage of class V composite resins was also compared with that of scanning electron microscopy (SEM). Methods Class V cavities were prepared on the buccal surface of sixteen extracted caries-free human premolars. Cavities were conditioned and filled with composite resin. The teeth were immersed in 50% weight/weight aqueous silver nitrate solution for 24 hours. They were then taken out and rinsed with distilled water. Next, they were put in a developing solution. They were first viewed with New Tom and Planmeca CBCT units and were then sectioned and evaluated by an SEM. Results The results of the Wilcoxon signed-rank test showed no significant difference between the mean marginal leakage scores of New Tom and Planmeca CBCT images (p value = 0.157) and between those of New Tom CBCT and SEM images (p value = 0.098). However, there was a significant difference between the mean marginal leakage scores of Planmeca CBCT and SEM images (p value = 0.023). Conclusion There were no significant differences between New Tom and Planmeca CBCT units in the detection of marginal leakage of class V composite resins. However, when these CBCT units were compared with the SEM, the New Tom CBCT unit could detect the marginal leakage better than Planmeca.


Introduction
Marginal leakage in composite resin restorations results in undesirable consequences such as tooth sensitivity, recurrent caries, restoration margin discoloration, pulpal irritation, and eventually restoration failure [1][2][3]. Among different composite resin restorations, class V cavities have a relatively small configuration factor; consequently, the composite resins have less effective mechanical properties, and the bonding strength of the adhesive defines the outcome of the restoration. Hence, insufficient bond in class V composite resin cavities results in marginal leakage [1].
Various in vitro methods have been used to investigate the marginal leakage of composite resin restorations. ese methods involve the use of biological, chemical, electrical, physical, and radioactive components. Among them, scanning electron microscopy (SEM) has proved to provide a better direct observation of the adaptation of a restorative material to the cavity margin [2]. Dye penetration methods are based on immersing the tooth sample in a dye solution and cutting the tooth through the center of the restoration to visualize the leakage of the restoration margins under a stereo microscope or an SEM. One disadvantage of these techniques is the loss of tooth structure integrity because of the need to cut the tooth to observe the leakage. erefore, nondestructive radiographic methods have been introduced for the assessment of leakage. e introduction of 3D radiographic methods with a small voxel size has helped the nondestructive detection of marginal leakage along the restoration margins. Two techniques have already been applied in this regard. e first one is microcomputed tomography (micro-CT), which is suggested for the assessment of marginal leakage in adhesive restorations and pit and fissure sealants. Another 3D technique in the field of dentistry is cone-beam computed tomography (CBCT), which has previously been used to detect small details like recurrent caries and vertical root fractures. It has recently been shown that CBCT can detect the marginal leakage of class V composite resin restorations in comparison to dye penetration methods [1,[4][5][6][7][8].
e increasing use of CBCT in dentistry to study fine details is undeniable.
ere are various CBCT units with different features for diagnostic purposes. None of previous studies has compared the ability of different brands of CBCT units to detect the marginal leakage of class V composite resin restorations. Furthermore, none of them has compared CBCT with SEM regarding their ability to detect the marginal leakage of class V composite resin restorations, which is the gold standard.
ere are different brands of CBCT devices around the world. However, in the geographical area where the study was conducted, only two brands (New Tom and Planmeca) were mostly available. is study was designed to assess the ability of two brands of CBCT to detect the marginal leakage of class V composite resin restorations. e ability of each of the two brands of CBCT to detect the marginal leakage of class V composite resins was also compared with that of SEM.

Sample Selection.
Sixteen caries-free human premolars were obtained from the patients who had their teeth extracted for orthodontic purposes. Patients' informed consent was obtained before the collection of samples. e teeth were cleaned of calculus, soft tissue, and debris by hand instrumentation after extraction. en, they were stored in 0.5% chloramine-T solution in the refrigerator (4°C). It took three months to collect the extracted teeth. All the teeth were prepared and restored by a single operator to prevent interoperator bias.

Sample Preparation.
A class V cavity was prepared on the buccal surface of each tooth by a cylindrical diamond bur (Tizkavan, Iran) mounted on a high-speed headpiece using air-water coolant spray (CH-4T5NSK B2/B3, Japan A1101800). Every four cavities were prepared by a new bur to avoid crack formation at the edges of the cavity. e cavities were checked by using a stereo microscope (Trinocular Zoom Stereo Microscope, SMP 200; HP, USA) to find probable cracks in the margin of cavities. e marginally cracked teeth were eliminated from the study to avoid false positive results due to the permeation of tracer through these spaces.
Class V cavities were 3 mm in width, 1.5 mm in depth, and 2.5 mm in height. e cavities were so prepared that the gingival margin of each cavity was extended beyond the cementoenamel junction (CEJ) onto the cementum, and the occlusal surface was limited to the enamel wall. e occlusal and gingival cavosurface margins of cavities were sharp and nonbeveled.

Composite Resin Filling.
e cavities were conditioned with the Clearfil SE Bond adhesive system (SE: self-etching) (lot no. 71167; Kuraray Medical Inc., Okayama, Japan). e adhesive system was applied according to the manufacturers' instructions. e teeth were restored with a composite resin (GRADIA DIRECT, Anterior, CE0086, Japan). e composite restorative material was applied and condensed incrementally until the preparations were completely filled. Each increment was light-polymerized for 40 seconds prior to the placement of the subsequent increment using an LED light cure with 700 mW/cm 2 light intensity (LED Turbo light cure; Taiwan). en, the extra parts of restorations were removed by a polishing bur and disk.

Silver Nitrate Infiltration.
e restored teeth were kept in distilled water for 24 hours at 37°C. To prevent the penetration of silver into areas other than the exposed margins, the tooth surfaces were sealed with two layers of nail varnish to be within approximately 1 mm of the restoration margins. e teeth were left at room temperature for one day to allow the nail varnish to dry. en, they were immersed in 50% weight/weight aqueous silver nitrate solution for 24 hours in a dark place at room temperature. After that, the teeth were extracted and rinsed with distilled water. Next, they were put into a developing solution (Dental X-Ray Developer, KONIX; TURKUAZ, ISTANBUL, TURKEY) for eight hours and were then removed and abundantly rinsed with water. e teeth were cleaned using a toothbrush to remove silver depositions on the surfaces.

CBCT Imaging.
e teeth were mounted in the sockets of a human mandible bone and observed by CBCT. CBCT images were taken by two units, including New Tom CBCT unit (New Tom GiANO; Quantitative Radiology, Verona, Italy) and Planmeca CBCT unit (Planmeca ProMax 3D Mid machine; Helsinki, Finland) ( Figure 1). e adjusted scan parameters of the New Tom CBCT unit were 90 kV, 3 mA, and 9 seconds. e field of view (FOV) was 8 * 11 cm Hi-Res., and the voxel size was 75 µm. e adjusted scan parameters of the Planmeca CBCT unit were 90 kV, 12.5 mA, and 15 seconds. e FOV was 10 * 10 cm, and the voxel size was 150 µm. e projections were analyzed by NNT (New Tom software) and Planmeca Romexis software in the crosssectional plane (slice thickness � 0.5 mm and slice distance � 0.1 mm) by two maxillofacial radiologists to detect marginal leakage along the gingival and coronal walls of class V cavities. e observers were blind to each other's responses. No software was used for artifact reduction.  [9]: 0: no leakage visible I: penetration of dye less than half of the gingival floor depth II: penetration of dye more than half of the gingival or occlusal floor depth without reaching the axial angle III: penetration of dye along the axial wall and toward the pulp

Statistical Analysis.
e interobserver agreement was determined with Cohen's kappa coefficient. Because two observers analyzed every tooth in each method, the average response of the two observers was taken for comparisons. Statistical comparisons were made between the two brands of CBCT units and then between each CBCT unit and SEM to determine the leakage scores of each method. Since the data distribution was not normal, the Wilcoxon signed-rank test was used to compare the gingival wall leakage scores between the above two dependent groups. e data were analyzed by the Statistical Package for Social Sciences (SPSS) (version 22; SPSS Inc., Chicago, IL). e significance level was set at 0.05.

Results
e leakage scores of restorations in the gingival floor measured by each technique are presented in Table 1. Assessment of interobserver agreement for New Tom CBCT and Planmeca CBCT images showed a Cohen's kappa coefficient of 0.887 (p value � 0.001), and assessment of interobserver agreement for SEM images showed a Cohen's kappa coefficient of 1.00 (p value � 0.00).
e results of the Wilcoxon signed-rank test showed no significant difference between the mean marginal leakage scores of New Tom and Planmeca CBCT images (p value � 0.157) and between the mean marginal leakage scores of New Tom CBCT and SEM images (p value � 0.098). However, the results of the Wilcoxon signed-rank test showed a significant difference between the mean marginal leakage scores of Planmeca CBCTand SEM images (p value � 0.023), as the marginal leakage scores of Planmeca CBCT were higher than those of SEM images (Tables 2-4). Both observers reported the same scores (0: no leakage visible) for the coronal walls in CBCT and SEM examinations. Figures 2-4 show the marginal leakage along the gingival and coronal walls of class V composite resin restorations using SEM and CBCT units.

Discussion
Marginal leakage is an adverse outcome caused by insufficient bond of composite resin restorations. is study was performed to assess the ability of two brands of CBCT units to detect the marginal leakage of class V composite resin restorations in comparison with SEM. e results showed that CBCT could detect the marginal leakage of class V composite resin restorations. Although there were no statistically significant differences between the two brands of CBCT in the detection of marginal leakage, the comparison of each of these two brands with SEM showed different results. ere was no significant difference between the marginal leakage scores of New Tom CBCT and SEM images, while there was a significant difference between the marginal leakage scores of Planmeca CBCTand SEM images. Indeed, the New Tom CBCT unit had better performance than the Planmeca CBCT unit in detecting marginal leakage.
is better performance of the New Tom CBCTunit could be related to its smaller field of view (FOV) and the smaller voxel size. A smaller FOV reduces the scatter radiation, which degrades the image quality. e scatter radiation is recorded by the cone-beam area detector. is additional recording of X-ray beam is called noise. erefore, the selection of a smaller FOV reduces the image noise and enhances its quality. Moreover, it has been proven that the presentation of details on a CBCT image strongly depends on spatial resolution (SR). SR is determined by the voxel size, and a smaller voxel size causes higher SR [10].
In this study, the smallest FOV was selected in both CBCT examinations to enhance the ability of observers to Our results showed that even this small difference in the FOV between the two units affected the amount of noise and degraded the image quality. In other words, the larger FOV in the Planmeca CBCT unit was associated with more noise production, which reduced the image quality. Moreover, the New Tom CBCT unit had a smaller voxel size and higher SR than the Planmeca unit. e minimum voxel size was fixed in the two devices (75 µm in New Tom and 150 µm in Planmeca), and it was not possible to equalize the voxel size of both devices in this study. erefore, the combination of higher SR and smaller FOV contributed to the better performance of the New Tom CBCT unit than Planmeca in the detection of marginal leakage.
is study assessed the marginal leakage of both coronal and gingival walls of class V composite resin restorations. None of the tooth samples showed marginal leakage on the  Observer 2  SEM  1  1  1  1  1  0  0  2  0  0  1  1  0  0  3  2  2  2  2  2     4 e Scientific World Journal coronal walls of class V composite resin restorations using SEM and CBCT. e coronal wall contains enamel, and the bond strength to enamel is more than that of dentin. Consequently, the marginal leakage of the coronal wall is less than that of the gingival floor. e results of this study were in agreement with those of other studies that used dye penetration and micro-CT methods to assess marginal leakage along coronal and gingival walls [1-4, 11, 12]. CBCT has limitations in assessing the marginal leakage of the coronal wall of class V cavities. As shown in Figure 5, SEM can reveal the marginal leakage of the coronal walls as well as the gingival walls of class V composite resin restorations. However, this might not be always true for the CBCT because some parts of the coronal wall of class V cavities are located in the enamel, and the radiopacity of enamel obscures the radiopacity of silver nitrate tracer. As a result, the CBCT observer may underestimate the amount of leakage on the coronal wall. However, when the leakage extends to the dentin beyond the enamel, it is definitively recognizable on the CBCT images. is can be considered one of the limitations of CBCT against SEM in the detection of marginal leakage. e Scientific World Journal 5 We could not find any similar articles in the literature that compared the capability of CBCT and SEM in the detection of marginal leakage. e most similar one was the study of Khoroushi et al. which compared the ability of CBCT (New Tom VGi Evo) and stereo microscopes to detect the marginal leakage of class V composite resin restorations and found no significant difference between these two methods [5]. e present study is basically different from the above study in that it compared the ability of two CBCTunits (NewTom GiANO and Planmeca) and SEM to detect marginal leakage. Since the use of stereo microscope for the detection of marginal leakage has become obsolete and has now been replaced by SEM, it seems necessary to compare the ability of CBCT and SEM to detect marginal leakage.
As shown in Rengo et al. study on micro-CT, one of the advantages of SEM over CT is the ability of SEM in image magnification [1]. is study used 80-250x magnification to assess the leakage through SEM images. e entire toothrestoration interface was quite visible at 80x magnification, and as the SEM magnification was increased above 80x, more details became visible and the silver nitrate tracing became more accurate ( Figure 6). However, increasing the magnification of CBCT images degrades the ability to evaluate the details due to the loss of image clarity. is can be taken into account as another limitation of CBCT in the detection of leakage.
As a previous study has shown, the proximity of the radiopacity of the tracer to the gingival wall of the cavity can mislead the observers to underestimate the marginal leakage scores on the CBCT images [5]. In the present study, the observers who evaluated the CBCT images encountered this problem, too. However, SEM did not show such a limitation because zooming SEM images could depict this proximity very well and prevented misinterpretation. Nevertheless, one advantage of CBCT over SEM is being nondestructive for tooth integrity. CBCT makes the evaluation of margins possible without tooth sectioning. Another advantage of CBCT is the ability to display the multiplanar images of the restoration margins and to provide the possibility of moving between sectional images in any plane along the restoration margin [5,13].
Silver nitrate solution was used in this study as a tracer because of the chemical and radiographic properties of silver in leakage examinations. ere is no agreement on the ideal concentration of silver nitrate and the optimal time of tooth immersion for CT observations. Yet, it appears that tooth immersion in 50% silver nitrate solution for 14-24 hours provides appropriate interface staining of silver ions and displays a good radiopaque contrast [1,4,14,15].
One of the limitations of the present study was the lack of soft tissue simulation. Soft tissue produces some noises in the CBCT images, which degrades the image quality and may affect the interpretations [10]. Another limitation of this study was comparison of only the two brands NewTom and Planmeca for reasons beyond the choice of researchers. Considering the importance of FOV and voxel size in the detection of marginal leakage and wide variations of such components in different brands of CBCT, future studies are suggested to compare other CBCT brands with SEM.

Conclusion
Within the limitations of this study, it can be concluded that there were no significant differences between New Tom and Planmeca CBCT units in detecting the marginal leakage of class V composite resins. However, when these CBCT units were compared with the SEM, the New Tom CBCT unit could detect the marginal leakage better than Planmeca.

SEM:
Scanning electron microscopy Micro-CT: Microcomputed tomography CBCT: Cone-beam computed tomography CT: Computed tomography SR: Spatial resolution USA: United States of America CEJ: Cementoenamel junction 3D: ree-dimensional C: Degree Celsius mm: Millimeter cm: Centimeter µm: Micrometer SE: Self-etching LED: Light-emitting diode NSK: Machinery manufacturing company mA: Milliampere kV: Kilovoltage mW: Megawatt FOV: Field of view Hi-Res: High resolution NNT: New Tom software Inc.: Incorporated CO: Company or corporation SPSS: Statistical Package for Social Sciences IL: Illinois p value: Calculated probability.
Data Availability e data used in the analysis of the present study can be provided by the corresponding author upon request. Disclosure e paper, or any part of it, has not been submitted or published and will not be submitted elsewhere for publication while being considered by the journal.