Abstract
Objectives:
To assess the distortion of high-density materials using two CBCT devices presenting convex triangular and cylindrical fields of view (FOVs).
Methods and materials:
Four high-density cylinders were individually placed in a polymethylmethacrylate phantom. 192 CBCT scans were acquired using the convex triangular and cylindrical FOVs of Veraviewepocs® R100 (R100) and Veraview® X800 (X800) devices. Using HorosTM’s software, two oral radiologists determined the cylinders’ horizontal and vertical dimensional alterations. Nine oral radiologists subjectively identified each cylinder’s axial shape distortion. Statistical analysis comprised Multiway ANOVA (α = 5%), and the Kruskal–Wallis test.
Results:
The distortion in the axial plane was greater in the convex triangular FOVs for both devices in almost all the materials (p < 0.05). The evaluators subjectively identified a shape distortion in both FOVs for R100 device (p < 0.001), while no distortion was identified for X800 device (p = 0.620). A vertical magnification of all materials was observed in both FOVs for both devices (p < 0.05). No differences among vertical regions (p = 0.988) nor FOVs (p = 0.544) were found for the R100 device, while all materials showed higher magnification in all regions in the cylindrical FOV (p < 0.001) of the X800 device.
Conclusions:
The convex triangular FOV influenced the axial distortion of the high-density materials in both devices. A vertical magnification was observed in both FOVs of both devices, but it was greater in the cylindrical FOV of the X800 device.
Keywords: Artifacts, Cone-Beam Computed Tomography, Dental Materials, Magnification
Introduction
The cone-beam computed tomography (CBCT) current market displays a wide variety of devices that offer different features aiming high image quality and dose reduction. Advances in operating systems, image detectors and reconstruction algorithms allow the acquisition and display of images with high resolution and a wide dynamic range. 1 Innovations in detectors, iterative reconstruction, beam collimation, optimized filtering, current modulation, and low-dose protocols allow maintaining image quality and decreasing patient dose. 2
According to the Veraviewepocs® R100 device manufacturer’s specifications, the convex triangular field of view (FOV) was conceived as an innovative way to diminish the radiation dose. It results from a variation of the 80-mm-diameter cylindrical FOV. Thus, a 180° eccentric trajectory (i.e. triangular path) of the radiation source during the scan acquisition decreases the intensity of the beam and, consequently, the dose. In addition, this modified trajectory results in a convex triangular-shaped FOV that adapts to the curved shape of the dental arches. 3,4
High-density and high atomic number materials generate artifacts in the reconstructed CBCT image, representing one of the drawbacks of CBCT since artifacts can reduce image quality. 5,6 The volumetric alteration artifact (also known as blooming) needs to be further assessed regarding its clinical impact, since this type of artifact interferes in the surrounding voxels of a high-density material object. 7 Furthermore, the dimensional analysis of high-density materials inside the FOV allows quantifying the expression of the volumetric alteration in terms of the object’s distortion. 8 This volumetric alteration may appear as a contraction (when the object’s dimensions are evenly reduced), a magnification (when the object’s dimensions are evenly increased), or a shape distortion (when the object’s dimensions change unevenly). 8 Moreover, knowing the influence of the FOV shape in the volumetric alteration artifact is relevant due to clinical implications when using high-density materials in the vicinity of thin anatomical structures such as the cortical bone and the radicular dentin. 9–13 Therefore, this in-vitro study aimed to objectively and subjectively assess the shape distortion of high-density materials commonly used in the Dentistry field, using two CBCT devices presenting both convex triangular and cylindrical FOVs.
Methods and materials
Sample preparation
Four high-density materials were used to manufacture standardized cylinders: chromium–cobalt (Scardua Laboratory, Vila Velha, Brazil), gutta-percha (Dentsply, Petrópolis, Rio de Janeiro, Brazil), titanium (S.I.N. Implantes, São Paulo, Brazil), and zirconium (Scardua Laboratory, Vila Velha, Brazil). The chromium–cobalt, titanium and zirconium cylinders were fabricated using a computer-aided design and manufacturing system (CAD-CAM), while the gutta-percha cylinder was fabricated using a silicone matrix following the manufacturer’s instructions. Once manufactured, the physical dimensions of each cylinder were measured three times using a digital caliper to guarantee the measurements’ accuracy (Table 1).
Table 1.
Dimensions of the cylinders according to each high-density material
| Material | Height (mm) | Diameter (mm) | Radius (mm) |
|---|---|---|---|
| Chromium–cobalt | 5.50 | 5.40 | 2.70 |
| Gutta-percha | 5.50 | 5.50 | 2.75 |
| Titanium | 5.50 | 5.50 | 2.75 |
| Zirconium | 5.50 | 5.40 | 2.70 |
Scan acquisition
Each high-density cylinder was placed in a 100-mm-diameter and 43.5-mm-height polymethylmethacrylate (PMMA) phantom with three 14.5-mm-height layers each (Figure 1). 14 The middle and the superior layers had eight holes each, arranged in a lower dental arch shape, while the inferior layer had no hole and served as a support. One additional hole was present on each side of each layer; these served for the stabilization of the three layers by inserting 43.5-mm-height PMMA cylinders inside. Only one high-density cylinder was used during each CBCT scan to prevent the possible intervention of artifacts arising from other high-density cylinders. The high-density cylinder was inserted in the middle layer and, since it did not occupy the full height of the hole, the empty spaces were filled with PMMA cylinders. The adjacent empty holes were also filled with PMMA cylinders for acquisitions. Four positions, considering different clinic-like variations, were selected to place the high-density materials: left-anterior, right-anterior, left-posterior, and right-posterior. The manufacturer quality control platforms and laser beams were used to standardize the phantom positioning in each CBCT device (Figure 1).
Figure 1.
(1a) Phantom inside the X800 CBCT device. (1b) Frontal and superior views of the phantom showing a high-density cylinder in the left-anterior position (continuous line). CBCT, cone-beam CT; LA, left-anterior; LP, left-posterior; RA, right anterior; RP, right-posterior.
Tomographic scans of the cylinders were acquired using Veraviewepocs® R100 (R100) and Veraview® X800 (X800) (J. Morita Mfg. Corp., Kyoto, Japan) CBCT devices. Two FOV shapes were used in both CBCT devices: cylindrical (80 diameter x 80 height mm) and convex triangular (100 diameter equivalent x 80 height mm). Exposure parameters were standardized at 90 kVp, 5 mA, and 9.4 s; except the voxel size, which was automatically determined by the devices for each FOV used. Thus, the voxel size of the cylindrical FOVs of both devices, and of the convex triangular FOV of the X800 device was 125 µm, while in the convex triangular FOV of the R100 device it was 160 µm. This choice was made to standardize the FOVs vertical sizes since it was not possible to acquire an 80 mm height triangular FOV with the 125 µm voxel size. Each experimental condition was repeated three times, resulting in a total of 192 CBCT acquisitions. The scans were exported in the DICOM (Digital Imaging and Communications in Medicine) format and randomized for further objective and subjective assessments.
Image evaluation
Two oral radiologists with 3 years of experience in the assessment of CBCT scans, blinded for the experimental conditions and calibrated for the objective analysis, performed image evaluation using the free and open-source medical image viewer software HorosTM v. 3.3.6 (Nimble Co LLC Purview, Annapolis, MD). Brightness and contrast adjustments were not allowed since a previous study observed the influence of windowing adjustments in the volumetric alteration artifact. 15 One of the evaluators assessed the whole sample for the objective analysis and reassessed 25% of it after 30 days, to calculate the intraexaminer reproducibility. The second evaluator assessed 25% of the sample to calculate interexaminer reproducibility, validating the first evaluator as the main evaluator. For image assessment, the evaluators determined the regions of linear measurements in the axial, coronal, and sagittal reconstructions.
For the measurements in the axial reconstruction, the theoretical geometric center of each cylinder’s image (distorted or not) was determined. For this, the evaluators draw a rectangle limiting the visible perimeter of the cylinder. Posteriorly, they drew two lines crossing the rectangle from the vertices of it. The point where the lines intersected was considered as the theoretical geometric center of the cylinder (Figure 2a). To assess the horizontal shape distortion, the evaluators measured the distance between the theoretical geometric center and the border of each cylinder in eight directions at specific angles for better reproducibility of the measurements (0°, left; 45°, left-anterior; 90°, anterior; 135°, right-anterior; 180°, right; 225°, right-posterior; 270°, posterior; and 315°, left-posterior). This process was done at three levels (superior, middle, and inferior) of each cylinder’s axial reconstructions (Figure 3a). As the CBCT scans were made so that each cylinder was in the middle height of the FOV, the middle of the cylinder was determined by picking the central axial reconstruction. For the R100 device, the superior and inferior axial reconstructions were placed at 16 axial reconstructions, which equals 2.56 mm (16 × 0.160 mm voxel). Similarly, for the X800 device, the superior and inferior axial reconstructions were placed at 20 axial reconstructions, which equals 2.50 mm (20 × 0.125 mm voxel).
Figure 2.
Determination of the measurement sites. (a) Theoretical geometric center of the cylinder in the axial reconstruction. (b) Center of the cylinder in the coronal and sagittal reconstructions. A, anterior; L, left; P, posterior; R, Right.
Figure 3.
Measurement regions. (a) Axial reconstructions. (b) Coronal and sagittal reconstructions
The center in the coronal and the sagittal reconstructions was determined by center of the cylinder previously defined in the axial reconstruction. The anterior, posterior, right, and left measurements were taken 2.5 mm from the center, with a safety margin of 0.5 mm from the border (Figure 2b). Thus, to assess the vertical shape distortion of the cylinders, the evaluators measured the height of the cylinders at five different regions (right, central, left, anterior, and posterior) using the coronal and sagittal reconstructions (Figure 3b).
For the subjective analysis, nine oral radiologists with at least 3 years of experience and unaware of the factors under study evaluated the images. The authors recorded 10-s videos of the axial reconstructions of all CBCT scans, using a 4x zoom, so the evaluators could dynamically visualize the whole cylinder without identifying the FOV shape (cylindrical or convex triangular), material, device or protocol used. After assessing each video of each experimental condition, the evaluators had to determine the cylinder’s shape in the axial reconstruction as convex triangular, intermediary (nor quite a convex triangle nor quite a circle), or circular.
Statistical analysis
The measurements made in the tomographic images were expressed in millimeters and were compared to the physical measurements to determine the shape distortion. Statistics comprised intraclass correlation coefficient (ICC) for intra- and interevaluator reproducibility, one-sample t-test to confirm the alteration of the dimensions of each cylinder (α = 5%), and multiway ANOVA (α = 5%) to determine the sites of distortion. For the subjective evaluation, weighted κ was used to determine the intra- and interevaluators reproducibility, and the Kruskal–Wallis with Dunn post-hoc tests assessed the differences between the cylinder’s image shape in the axial reconstruction (convex triangular, intermediary, circular).
Results
The ICC showed intra- and interevaluator reproducibility ranging from almost perfect (0.969) to perfect (1.00) for the axial measurements, and almost perfect (0.996 and 0.980) for the vertical measurements. For the subjective evaluations, the weighted κ test showed intraevaluator reproducibility ranging from substantial (0.694) to almost perfect (0.984), and interevaluator reproducibility ranging from fair (0.391) to substantial (0.799) (Table 2).
Table 2.
Weighted κ test for intra- and interevaluator reproducibility concerning the subjective analysis
| Evaluators | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |||
| Evaluators | 1 | 0.892 | 0.671 | 0.571 | 0.658 | 0.799 | 0.754 | 0.672 | 0.767 | 0.658 | |
| 2 | 0.694 | 0.478 | 0.610 | 0.634 | 0.577 | 0.680 | 0.666 | 0.701 | |||
| 3 | 0.877 | 0.391 | 0.572 | 0.581 | 0.575 | 0.603 | 0.551 | ||||
| 4 | 0.738 | 0.697 | 0.624 | 0.538 | 0.596 | 0.523 | |||||
| 5 | 0.805 | 0.763 | 0.661 | 0.735 | 0.648 | ||||||
| 6 | 0.872 | 0.617 | 0.652 | 0.546 | |||||||
| 7 | 0.872 | 0.733 | 0.710 | ||||||||
| 8 | 0.760 | 0.776 | |||||||||
| 9 | 0.984 | ||||||||||
Diagonal comparisons indicate intraevaluator reproducibilities.
The one-sample t-test confirmed the vertical and horizontal dimensional alterations in all materials for both devices (p < 0.05). The average radius in the axial reconstructions ranged from decreasing in most directions to increasing in some directions. Conversely, the average heights in all regions were higher than those of the physical cylinders for both FOVs and devices.
According to multiway ANOVA, for radii measurements in the R100 device, no significant differences were found among the levels assessed in all the materials (p > 0.05). For the cylindrical FOV, in all materials, the measurements in the left-anterior and left-posterior directions were the closest to the physical measurements. In all other directions, the materials showed contraction (p < 0.05), being more pronounced in the right-posterior and right-anterior directions (p < 0.05). For the convex triangular FOV, measurements from the left-posterior direction were closer to the real dimensions in all materials. The left-anterior direction presented magnification (p < 0.05), and all other directions presented contraction, being more significant in the right-posterior and right-anterior directions. It is also observed that this pattern of contraction/magnification in the directions mentioned above was more pronounced for the convex triangular FOV than the cylindrical FOV of the R100 device (p < 0.05) (Figure 4).
Figure 4.
Levels’ average radii in mm (standard deviation) of the cylinders according to the evaluated directions and fields of view in the Veraviewepocs® R100) device. CoCr (Chromium–cobalt), (Gu) (Gutta-percha), (Ti) (Titanium), (Zr) (Zirconium). Continuous line (border of the physical cylinder). Discontinuous line (border of the cylinder in the CBCT scan). Lowercase letters show statistical differences among directions. Asterisk (*) shows differences between fields of view (considering the same direction and material), according to multiway ANOVA. CBCT, cone-beam CT; FOV, field of view.
Regarding the radii measurements in the X800 device, no difference among the levels was observed (p > 0.05). In general, for the cylindrical FOV, few statistically significant differences were observed among the directions. However, it is possible to observe that the contraction of all materials was a trend. Additionally, gutta-percha and zirconium contracted unevenly, demonstrating a shape distortion (p < 0.05), while chromium–cobalt and titanium contracted evenly without distorting (p > 0.05). For the convex triangular FOV, chromium–cobalt and titanium showed greater contraction in the right-posterior and left-posterior directions (p < 0.05). Zirconium also showed similar behavior; however, the contraction was also observed in the right, left, and anterior directions. Gutta-percha showed greater contraction in all directions, except in the left-anterior and right-anterior directions (p < 0.05) (Figure 5).
Figure 5.
Levels’ average radii in mm (standard deviation) of the cylinders according to the evaluated directions and fields of view in the Veraviewepocs® X800) device. CoCr (Chromium–cobalt), (Gu) (Gutta-percha), (Ti) (Titanium), (Zr) (Zirconium). Continuous line (border of the physical cylinder). Discontinuous line (border of the cylinder in the CBCT scan). Lowercase letters show statistical differences among directions. Asterisk (*) shows differences between fields of view (considering the same direction and material), according to multiway ANOVA. CBCT, cone-beam CT; FOV, field of view.
Regarding the measurements performed in the coronal and sagittal reconstructions of the cylinder, no statistically significant differences were found among regions (p > 0.05) nor between FOVs (p > 0.05) in the R100 device, which indicates a magnification of all materials. On the other hand, all materials showed higher magnification in all regions in the cylindrical FOV (p < 0.001) than in the convex triangular FOV of the X800 device (Table 3).
Table 3.
Height in mm (standard deviation) of the cylinders according to the evaluated regions and fields of view in the Veraviewepocs® R100 and Veraview® X800 devices
| Material | Region | Field of view R100 | Field of view X800 | ||||
|---|---|---|---|---|---|---|---|
| Cylindrical | Convex triangular | p-value a | Cylindrical | Convex triangular | p-value a | ||
| Chromium–cobalt | Central | 5.92 (0.03) | 5.91 (0.03) | p = 0.988 | 6.09 (0.16) | 5.78 (0.04) | p = 0.139 |
| Right | 5.92 (0.03) | 5.92 (0.04) | 6.07 (0.16) | 5.70 (0.05) | |||
| Left | 5.93 (0.02) | 5.91 (0.02) | 6.08 (0.14) | 5.68 (0.06) | |||
| Anterior | 5.93 (0.03) | 5.92 (0.04) | 6.10 (0.15) | 5.69 (0.07) | |||
| Posterior | 5.95 (0.03) | 5.94 (0.03) | 6.09 (0.15) | 5.72 (0.08) | |||
| Gutta-percha | Central | 5.94 (0.03) | 5.92 (0.09) | 6.01 (0.04) | 5.69 (0.10) | ||
| Right | 5.93 (0.03) | 5.89 (0.10) | 6.00 (0.04) | 5.63 (0.12) | |||
| Left | 5.95 (0.03) | 5.92 (0.08) | 6.06 (0.11) | 5.63 (0.06) | |||
| Anterior | 5.92 (0.04) | 5.91 (0.02) | 6.01 (0.03) | 5.62 (0.08) | |||
| Posterior | 5.96 (0.03) | 5.93 (0.03) | 6.05 (0.09) | 5.61 (0.04) | |||
| Titanium | Central | 5.93 (0.03) | 6.06 (0.25) | 6.08 (0.14) | 5.63 (0.07) | ||
| Right | 5.93 (0.03) | 6.06 (0.23) | 6.07 (0.14) | 5.66 (0.07) | |||
| Left | 5.92 (0.03) | 6.05 (0.24) | 6.08 (0.15) | 5.61 (0.06) | |||
| Anterior | 5.94 (0.04) | 6.06 (0.23) | 6.05 (0.12) | 5.61 (0.06) | |||
| Posterior | 5.92 (0.03) | 6.07 (0.24) | 6.05 (0.11) | 5.57 (0.04) | |||
| Zirconium | Central | 5.94 (0.02) | 5.92 (0.16) | 6.11 (0.15) | 5.78 (0.11) | ||
| Right | 5.94 (0.02) | 5.92 (0.16) | 6.08 (0.16) | 5.76 (0.08) | |||
| Left | 5.94 (0.02) | 5.93 (0.15) | 6.10 (0.16) | 5.75 (0.10) | |||
| Anterior | 5.94 (0.03) | 5.94 (0.14) | 6.08 (0.15) | 5.72 (0.10) | |||
| Posterior | 5.96 (0.02) | 5.93 (0.15) | 6.09 (0.16) | 5.71 (0.10) | |||
| p-value b | p = 0.544 | p < 0.001 | |||||
p-value a indicates comparisons among regions for each device.
p-value b indicates comparisons between intradevice FOVs.
According to multiway ANOVA.
Concerning the subjective analysis, the Kruskal–Wallis test showed that the evaluators identified a convex triangular distortion of the cylinders in the convex triangular FOV and an intermediary distortion in the cylindrical FOV of the R100 device (p < 0.05) (Table 4). The Dunn post-hoc test was applied to show the differences in the evaluators' perception among the materials observed in the R100 device. In this sense, the titanium and zirconium cylinders in the convex triangular FOV were perceived with the greatest convex triangular distortion (p < 0.05), and gutta-percha and titanium in the cylindrical FOV were perceived as intermediary distorted (p < 0.05) (Figure 4). Conversely, no shape distortion was observed (i.e. a circular shape was determined) among all the materials in the X800 device (p = 0.620) (Figure 5).
Table 4.
Sum of ranks, median, ranks mean, mean, and standard deviation of the subjective analysis
| Convex triangular FOV | Cylindrical FOV | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| CrCo | Gu | Ti | Zr | CrCo | Gu | Ti | Zr | ||
| R100 | ∑ Ranks | 47.00 | 47.00 | 42.00 | 42.00 | 81.00 | 88.00 | 91.00 | 79.00 |
| Median | 1.00 | 1.00 | 1.00a | 1.00a | 2.00 | 2.00b | 2.00b | 2.00 | |
| Ranks Mean | 3.98 | 3.98 | 3.50 | 3.50 | 6.75 | 7.33 | 7.58 | 6.58 | |
| Mean | 1.08 | 1.08 | 1.00 | 1.00 | 1.750 | 1.92 | 1.92 | 1.67 | |
| Standard deviation | 0.29 | 0.29 | 0.00 | 0.00 | 0.62 | 0.67 | 0.51 | 0.43 | |
| p value Kruskal–Wallis | < 0.05 | ||||||||
| X800 | ∑ Ranks | 56.00 | 56.00 | 61.00 | 53.50 | 75.50 | 70.50 | 75.50 | 75.50 |
| Median | 3.00 | 3.00 | 3.00 | 3.00 | 3.00 | 3.00 | 3.00 | 3.00 | |
| Ranks mean | 4.67 | 4.67 | 5.08 | 4.46 | 6.29 | 5.87 | 6.29 | 6.29 | |
| Mean | 2.67 | 2.67 | 2.75 | 2.58 | 3.00 | 2.92 | 3,00 | 3.00 | |
| Standard deviation | 0.49 | 0.49 | 0.45 | 0.67 | 0.00 | 0.29 | 0.00 | 0.00 | |
| p value Kruskal–Wallis | =0.6206 | ||||||||
FOV, field of view.
CrCo, Chromium–Cobalt; Gu, Gutta-percha; Ti, Titanium; Zr, Zirconium.
Bold p value indicates statistically significant difference, according to Kruskal–Wallis test.
Different lowercase letters indicate statistically differences between the FOV shape considering intradevice comparisons, according to Dunn post-hoc test for the R100 device.
Discussion
Our study showed that the FOV shape influences the shape distortion in the horizontal plane, which was more significant in the convex triangular than in the cylindrical FOVs for both devices, and greater in the R100 device than in the X800 device. Furthermore, this shape distortion mimics a convex triangular shape, strongly related, but not exclusive, to the convex triangular FOVs.
The authors' hypothesis for this shape distortion is that the expression of this artifact relies on a combination of variables interacting during the scan acquisition, misinterpreted by the image detector, and finally shown in the reconstructed scan. In this sense, we must consider the crucial role of the interaction between the beam hardening and the partial volume effects with high-density materials in the reconstruction of the object’s boundaries. 5,6,16,17 In addition, the interaction of a continuous exposure and the rotation speed with the source-to-object trajectory could also be considered to present a role. Also, all the acquired scans in both devices had a rotation angle of 180°, which means fewer basis images available to reconstruct the final volume, thus an increased chance of misrepresentation of the scanned object. 18 Furthermore, the image formation geometry differs between protocols: the center of rotation is isocentric in the cylindrical FOV, while it is eccentric in the convex triangular FOV. 3,4 Thus, the constant velocity of the X-ray source in a 180 degree rotation angle around the object combined with the momentaneous trajectory modification, the beam hardening, and the partial volume effect could all be responsible for the greater shape distortion shown in the convex triangular FOVs.
A slight shape distortion was observed in the axial reconstruction of almost all the materials in the cylindrical FOV for both devices, except for the titanium cylinder in the X800 device which did not present any distortion. These results partially agree with those observed in a previous study which identified mainly shape distortion followed by magnification and contraction of high-density materials. 8 No magnification in the axial plane of the cylindrical FOVs was observed, only shape distortion and contraction. The authors believe that these discrepancies could be associated to the interaction between the high-density materials with the reconstruction algorithms of each CBCT device.
Besides, it is interesting how titanium presents a more stable behavior than zirconium, which could be of clinical relevance for implant dentistry; this different behavior is probably related to the fact that titanium presents a lower atomic number than zirconium. Even more interesting is that this material showed an even contraction in the cylindrical FOV of the X800 device, contrary to the shape distortion observed in cylindrical FOV of the R100 device. This suggests that there are differences between the devices during the scan acquisition and reconstruction. In that sense, according to the manufacturer´s specifications, the R100 device presents a fixed source of radiation with a negative angulation of the X-ray beam, making it also capable of obtaining panoramic images. On the other hand, in the X800 device, also capable of acquiring panoramic images, the angle of the X-ray beam can be adjusted from horizontal (for CBCT exposures) to 5° raised for panoramic exposures by shifting the Flat Panel Detector.
When comparing the shape distortion among the three levels of all the materials in both FOVs of both devices, no differences were observed, suggesting that when the volumetric alteration artifact of a high-density material is present, the dimensional alterations of the object are homogeneously expressed. A previous study assessed the dimensional alteration of high-density materials through linear measurements considering only the center of the object. 8 However, to the best of the authors’ knowledge, no previous studies assessed the dimensional alteration in the axial plane concerning different object levels.
Contrary to the axial analysis, the vertical assessment of shape distortion clearly showed that all materials evenly augmented their mean height values when compared to the physical measurements in both devices and FOVs. In this sense, the homogeneous height augmentation could have a greater impact on the final volume of the reconstructed scan, even in the presence of regions of contraction in the axial reconstructions. Furthermore, this observation confirms what previous studies suggested about the segmenting difficulty of high-density cylinders' upper and lower boundaries. 4,19 This finding could be explained by the divergence of the X-rays beam that interacts with the sharp borders of the superior and inferior surfaces of the cylinder and creates two diffuse hyperdense regions at the top and bottom of it. This diffuse region was considered part of the volumetric alteration artifact in the present study as it couldn't allow the correct visualization of the adjacent structures.
The lack of difference in the magnitude of the height magnification in the R100 device suggests that even when there was a slight voxel size difference between the cylindrical and the convex triangular FOVs, the partial volume effect may not have an essential role in the height alteration. The role of the voxel size and the partial volume effect should be further investigated since this was not an objective of the present study. Contrary to this observation, the cylindrical FOV of the X800 device showed greater height mean values than the convex triangular FOV. This finding supports that the image formation geometry influences the expression of the volumetric alteration artifact and, consequently, the dimensional alteration of high-density materials. The possible reason for the minor expression in the convex triangular FOV is that with the source of radiation getting far from the object during the acquisition, the divergence of the X-ray beam temporarily decreases and, thus, the diffuse zone at the top and bottom of the high-density cylinder also decreases.
The subjective evaluation of the axial shape distortion showed that the R100 device presented a perceptible convex triangular distortion among all materials in the convex triangular FOV, especially for gutta-percha and titanium cylinders, while observing an intermediary shape distortion among all materials in the cylindrical FOV. This subjective observation agrees with the objective analysis, confirming that the distortion in the R100 device is relevant at the point to be detected by the evaluators. On the other hand, the lack of difference in the subjective evaluation for the X800 device contrasts with the objective assessment, which showed a contraction or just a slight distortion in the cylindrical FOV and a slight distortion in the convex triangular FOV. This observation calls attention to the evaluator’s difficulty in recognizing the high-density materials contractions or discrete distortions, which could explain the fair to substantial interevaluator reproducibility values.
Limitations and future studies
It is of relevance to state that even though the volumetric alteration artifact was assessed in this study, the volume of the objects was not measured, instead, their dimensions were. In order to achieve the objectives of the present study, a linear analysis was perceived as suitable. Thus, the authors found the theoretical geometric center by locating the point where the lines intersect at the corners of the superimposed rectangle to the CBCT cylinder´s image, ensuring that the process can be repeated accurately. Since there was a visual evident distortion, we intended to set a center for the cylinder´s image to objectively analyze the image distortion and make reliable comparisons between the different high-density image cylinders.
Attention should be given to the voxel size difference between the cylindrical FOV (125 µm) and the convex triangular FOV (160 µm) in the R100 device to standardize the FOV size and avoid its possible influence in the volumetric alteration artifact. Anyhow, this voxel size difference was just about 35 µm, and the authors believe that the artifact expression differences between FOVs is strongly influenced by the image geometry formation. In this sense, the fully standardization of the parameters in the X800 device confirmed that the image shape distortion is influenced by the image geometry formation. For a better understanding of the actual influence of voxel size on the dimensions and shape distortion of high-density materials, future studies should be carried out.
As an in vitro study, our findings should guide further clinical research but cannot be directly translated to the clinical scenario. Furthermore, it was assessed two CBCT devices with a specific FOV, and it doesn’t necessarily apply to other devices because of inner differences. In this sense, future ex vivo studies assessing dental materials in clinic-like conditions should be carried out to understand the clinical relevance of the high-density materials’ shape distortion.
Conclusions
The convex triangular FOV objectively influenced the shape distortion of the high-density materials in both devices among all materials in the axial plane. The shape distortion was detected among all the materials in the cylindrical FOV of the R100 device, while the contraction of all materials was observed in the cylindrical FOV of the X800 device. Furthermore, gutta-percha and zirconium contracted unevenly, demonstrating shape distortion, while titanium and chromium–cobalt showed an even contraction. It was observed a significant vertical magnification in both FOVs of both devices. Still, there was a greater magnification in the cylindrical FOV than in the convex triangular FOV of the X800 device. The evaluators subjectively identified a convex triangular distortion in the convex triangular FOV and an intermediary distortion in the cylindrical FOV of the R100 device. However, the evaluators did not recognize any distortion in both FOVs of the X800 device.
Footnotes
Acknowledgments: The first author is grateful to the University of Costa Rica for funding his postgraduate studies. Special thanks to Oral Radiology Section of the Faculty of Dentistry at the University of Costa Rica for their support during image acquisition with the Veravieraviewepocs R100 CBCT device, and to Dr. Ana Luisa Berrocal Domínguez for her support during image acquisition with the Veravieraview X800 CBCT device.
The authors thank S.I.N. Implantes for providing the titanium cylinders. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001.
Contributor Information
Deivi Cascante-Sequeira, Email: dantikes@gmail.com.
Fernanda Coelho-Silva, Email: silva.fernanda.coelho@gmail.com.
Lucas P Lopes Rosado, Email: lucaslopesrosado@gmail.com.
Luiza Valdemarca Lucca, Email: luiza_lucca@hotmail.com.
Deborah Queiroz Freitas, Email: deborahq@unicamp.br.
Sergio Lins de-Azevedo-Vaz, Email: sergiolinsv@gmail.com.
Francisco Haiter-Neto, Email: haiter@unicamp.br.
REFERENCES
- 1. Gaêta-Araujo H, Alzoubi T, Vasconcelos K de F, Orhan K, Pauwels R, Casselman JW, et al. Cone-beam computed tomography in dentomaxillofacial radiology: a two-decade overview. Dentomaxillofac Radiol 2020; 49(): 20200145. doi: 10.1259/dmfr.20200145 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Kiljunen T, Kaasalainen T, Suomalainen A, Kortesniemi M. Dental cone beam CT: A review. Phys Med 2015; 31: 844–60. doi: 10.1016/j.ejmp.2015.09.004 [DOI] [PubMed] [Google Scholar]
- 3. Rottke D, Dreger J, Sawada K, Honda K, Schulze D. Comparison of manual and dose reduction modes of a MORITA R100 CBCT. Dentomaxillofac Radiol 2019; 48: 20180009. doi: 10.1259/dmfr.20180009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cascante-Sequeira D, Coelho-Silva F, Rosado LPL, Freitas DQ, de-Azevedo-Vaz SL, Haiter-Neto F. Comparison of the expression of the volumetric alteration artifact in cylindrical and triangular fields of view in two cone-beam computed tomography devices. Clin Oral Investig 2022; 26: 1025–33. doi: 10.1007/s00784-021-04086-1 [DOI] [PubMed] [Google Scholar]
- 5. Schulze R, Heil U, Gross D, Bruellmann DD, Dranischnikow E, Schwanecke U, et al. Artefacts in CBCT: a review. Dentomaxillofac Radiol 2011; 40: 265–73. doi: 10.1259/dmfr/30642039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Pauwels R, Araki K, Siewerdsen JH, Thongvigitmanee SS. Technical aspects of dental CBCT: state of the art. Dentomaxillofac Radiol 2015; 44: 20140224. doi: 10.1259/dmfr.20140224 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Suetens P. X-ray computed tomography. In: Suetens P, ed. Fundamentals of medical imaging. 2nd ed. New York: Cambridge Univ Press; 2009, pp. 53–63. [Google Scholar]
- 8. Coelho-Silva F, Gaêta-Araujo H, Rosado LPL, Freitas DQ, Haiter-Neto F, de-Azevedo-Vaz SL. Distortion or magnification? an in Vitro cone-beam CT study of dimensional changes of objects with different compositions. Dentomaxillofac Radiol 2021; 50: 20210063. doi: 10.1259/dmfr.20210063 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Vanderstuyft T, Tarce M, Sanaan B, Jacobs R, de Faria Vasconcelos K, Quirynen M. Inaccuracy of buccal bone thickness estimation on cone-beam CT due to implant blooming: an ex-vivo study. J Clin Periodontol 2019; 46: 1134–43. doi: 10.1111/jcpe.13183 [DOI] [PubMed] [Google Scholar]
- 10. Gregoris Rabelo LE, Bueno MDR, Costa M da, de Musis CR, Estrela C de A, Guedes OA, et al. Blooming Artifact reduction using different cone-beam computed tomography software to analyze Endodontically treated teeth with Intracanal posts. Comput Biol Med 2021; 136: 104679. doi: 10.1016/j.compbiomed.2021.104679 [DOI] [PubMed] [Google Scholar]
- 11. Celikten B, Jacobs R, de Faria Vasconcelos K, Huang Y, Shaheen E, Nicolielo LFP, et al. Comparative evaluation of cone beam CT and micro-CT on blooming artifacts in human teeth filled with bioceramic sealers. Clin Oral Investig 2019; 23: 3267–73. doi: 10.1007/s00784-018-2748-8 [DOI] [PubMed] [Google Scholar]
- 12. Celikten B, Jacobs R, de Faria Vasconcelos K, Huang Y, Nicolielo LFP, Orhan K. Assessment of volumetric distortion artifact in filled root canals using different cone-beam computed tomographic devices. J Endod 2017; 43: 1517–21. doi: 10.1016/j.joen.2017.03.035 [DOI] [PubMed] [Google Scholar]
- 13. Rodrigues CT, Jacobs R, Vasconcelos KF, Lambrechts P, Rubira-Bullen IRF, Gaêta-Araujo H, et al. Infuence of CBCT-based volumetric distortion and beam hardening artefacts on the assessment of root canal flling quality in isthmus-containing molars. Dentomaxillofac Radiol 2021; 50(): 20200503. doi: 10.1259/dmfr.20200503 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Martins LAC, Queiroz PM, Nejaim Y, Vasconcelos K de F, Groppo FC, Haiter-Neto F. Evaluation of metal Artefacts for two CBCT devices with a new dental arch phantom. Dentomaxillofac Radiol 2020; 49: 20190385. doi: 10.1259/dmfr.20190385 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Coelho-Silva F, Martins LAC, Braga DA, Zandonade E, Haiter-Neto F, de-Azevedo-Vaz SL. Infuence of windowing and metal artefact reduction algorithms on the volumetric dimensions of fve Diferent high-density materials: a cone-beam CT study. Dentomaxillofac Radiol 2020; 49(): 20200039. doi: 10.1259/dmfr.20200039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Draenert FG, Coppenrath E, Herzog P, Müller S, Mueller-Lisse UG. Beam hardening artefacts occur in dental implant scans with the NewTom cone beam CT but not with the dental 4-row multidetector CT. Dentomaxillofac Radiol 2007; 36: 198–203. doi: 10.1259/dmfr/32579161 [DOI] [PubMed] [Google Scholar]
- 17. Schulze RKW, Berndt D, d’Hoedt B. On cone-beam computed tomography artifacts induced by titanium implants. Clin Oral Implants Res 2010; 21: 100–107. doi: 10.1111/j.1600-0501.2009.01817.x [DOI] [PubMed] [Google Scholar]
- 18. Scarfe WC, Farman AG. What is cone-beam CT and how does it work Dent Clin North Am 2008; 52: 707–30. doi: 10.1016/j.cden.2008.05.005 [DOI] [PubMed] [Google Scholar]
- 19. Codari M, de Faria Vasconcelos K, Ferreira Pinheiro Nicolielo L, Haiter Neto F, Jacobs R. Quantitative evaluation of metal artifacts using Diferent CBCT devices, high-density materials and Feld of views. Clin Oral Implants Res 2017; 28: 1509–14. doi: 10.1111/clr.13019 [DOI] [PubMed] [Google Scholar]