If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Reprint requests: Polyane Mazucatto Queiroz, DDS, MSc, Piracicaba Dental School, Department of Oral Diagnosis, Area of Oral Radiology, 901, Limeira Avenue, Piracicaba, Sao Paulo, Brazil 13414-903
The aim of this study was to evaluate the efficacy of a metal artifact reduction (MAR) algorithm in cone beam computed tomography (CBCT) images of dental materials obtained with different field-of-view (FOV) and voxel sizes.
Study Design
Two imaging phantoms were custom-made of acrylic resin. Each phantom had 3 cylinders made of the same dental material: dental amalgam or copper-aluminum alloy. CBCT scans were obtained separately for each of the imaging phantoms using the Picasso-Trio CBCT (Vatech, Hwaseong, Republic of Korea) unit at 4 FOV sizes and 2 voxel sizes. Each imaging phantom was scanned with and without MAR. All images were evaluated in the OnDemand3D software (Cybermed, Seoul, Republic of Korea) and image noise (gray value variability) was calculated as the standard deviation (SD) of the gray values of regions of interest around the dental material cylinders. Data were compared by the Friedman test and Dunn test (α = 0.05). Intraclass correlation coefficient (ICC) was calculated to assess intraobserver reliability.
Results
MAR significantly reduced (P < .05) image noise around the dental materials, irrespective of FOV and voxel sizes, with an ICC of 0.997.
Conclusions
The efficacy of MAR was similar for the different FOV and voxel sizes studied. Hence, imaging protocols and the use of MAR algorithm should be based on the selection criteria.
Clinical practice in association with diagnostic imaging should determine cone beam computed tomography (CBCT) exposure protocols based on the diagnostic task regardless of the use of a metal artifact reduction tool.
Numerous advantages of cone beam computed tomography (CBCT) over plain radiography have increased the acceptance of this imaging modality for some dental applications.
CBCT image artifact has shown to have a negative impact on the diagnostic process, and the presence of metallic materials in the oral cavity is among the most common causes of severe artifacts.
have been developed with the purpose of reducing potential CBCT image deterioration. Picasso Trio CBCT (Vatech, Hwaseong, Republic of Korea) is one of the CBCT units that make a MAR algorithm available. This is a post-processing tool that works during image reconstruction.
The presence of artifacts can compromise different diagnostic tasks such as the detection of root fractures in teeth with endodontic treatment and/or metal posts, and peri-implant evaluation.
Detection of periimplant fenestration and dehiscence with the use of two scan modes and the smallest voxel sizes of a cone-beam computed tomography device.
Oral Surg Oral Med Oral Pathol Oral Radiol.2013; 115: 121-127
Assessment of bucal marginal alveolar peri-implant and periodontal defects using a cone beam CT system with and without the application of metal artefact reduction mode.
Since CBCT imaging protocols may affect image quality and must be adjusted for each individual case, it is necessary to evaluate the effectiveness of MAR under these different conditions. Thus, the aim of this study was to evaluate the efficacy of a MAR algorithm in CBCT images of dental materials obtained with different field-of-view (FOV) and voxel sizes.
Materials and Methods
Imaging phantom preparation
Two imaging phantoms were custom-made of chemically activated acrylic resin (CAAR) (VIPI, Sao Paulo, Brazil) and 3 cylinder-shaped dental materials (diameter, 5.4 mm; height, 5.4 mm). One phantom contained 3 cylinders made of dental amalgam and the other contained 3 cylinders of copper-aluminum (Cu-Al) alloy.
CAAR was poured into a cylindrical polyvinyl chloride (PVC) pipe mold (Tigre, Sao Paulo, Brazil) with an internal diameter of 98 mm and a height of 40 mm. The dental material cylinders were vertically submerged 20 mm above the base and arranged in the vertices of an imaginary isosceles triangle (base, 58 mm; height, 39 mm).
CBCT scanning
Each phantom was positioned separately in the Picasso Trio CBCT unit (Vatech, Hwaseong, Republic of Korea) adjusted at 80 kVp, 3.7 mA, and acquisition time of 24 s. Eight scans were obtained with the combination of 4 FOV sizes (120 × 85 mm, 80 × 80 mm, 80 × 50 mm, 50 × 50 mm; diameter × height) and 2 voxel sizes (0.3 mm and 0.2 mm; Table I). The imaging phantom was centered in the middle of the FOV, except for the smallest FOV, in which each dental material cylinder was centered separately. All scans were repeated with the use of MAR. Examples of axial reconstructions of different FOV and voxel sizes used in this study are in Figures 1 and 2, respectively. Volumetric data were reconstructed in the native Ez3D software (E-WOO Technology, Seoul, Republic of Korea) and exported to DICOM file format.
Fig. 1Examples of axial reconstruction of cone beam computed tomography scans with a voxel size of 0.3 mm and different field-of-view (FOV) sizes: A and E, 120 × 85 mm; B and F, 80 × 80 mm; C and G, 80 × 50 mm; D and H, 50 × 50 mm. A-D and E-H are with and without metal artifact reduction (MAR), respectively.
Fig. 2Examples of axial reconstruction of cone beam computed tomography (CBCT) scans with a field-of-view (FOV) size of 80 × 80 mm and different voxel sizes: A and C, 0.3 mm; B and D, 0.2 mm. A, B and C, D are with and without metal artifact reduction (MAR), respectively.
All CBCT scans were imported to OnDemand3D software (Cybermed, Seoul, Republic of Korea) and a single examiner assessed all of them (Figures 1 and 2). On axial reconstructions, 6 circular regions of interest (ROIs) (diameter, 5.4 mm) were selected around and tangent to the borders of the middle height of the 3 dental material cylinders, comprising a total of 18 ROIs, as shown in Figure 3. Image noise (gray value variability) was calculated as the standard deviation (SD) of the gray values of the ROIs around the dental material cylinders. After 90 days, 50% of the sample was reevaluated to assess the reproducibility of the method.
Fig. 3Axial reconstruction of a cone beam computed tomography (CBCT) scan with 6 circular regions of interest (ROIs) located around each sample of dental material (dental amalgam), comprising a total of 18 ROIs.
All analyses were carried out using GraphPad Prism 6.0 (GraphPad Software, La Jolla, CA) and BioEstat 5.0 (Fundação Mamiraua, Belém, PA, Brazil). Intraclass correlation coefficient (ICC) was calculated to assess intraobserver reliability. Since the normality and homoscedasticity assumptions were not accomplished using the Kolmogorov-Smirnov test and Bartlett's test, data were compared by the Friedman test and Dunn test, with the significance level set at 5% (α = 0.05).
Results
The effect of MAR on image noise of CBCT images of dental materials obtained with different FOV and voxel sizes is shown in Table II and Figure 4. The use of MAR significantly reduced image noise of the ROIs around both the dental amalgam and Cu-Al alloy cylinders (P < .05), irrespective of FOV and voxel sizes, with an ICC of 0.997. In general, image noise was constant for all scanning parameters used in this study. Conversely, image noise of both dental materials differed from each other irrespective of scanning parameters, with lower values for Cu-Al alloy (Table III) when MAR was absent (P < .0001) or present (P < .0001).
Table IIMean values (SD) of SD gray value for different protocols and dental materials
FOV (mm)
Voxel (mm)
Dental amalgam
Cu-Al alloy
Without MAR
With MAR
Without MAR
With MAR
120 × 85
0.2
332.9 (±77.02)
113.0 (±40.3)
265.72 (±63.21)
105.2 (±35.1)
0.3
351.4 (±119.4)
125.0 (±46.16)
287.9 (±98.7)
125.7 (±44.3)
80 × 80
0.2
405.8 (±143.4)
103.6 (±31.6)
306.2 (±68.7)
101.0 (±32.9)
0.3
380.6 (±115.4)
132.2 (±46.4)
345.9 (±131.4)
124.1 (±45.7)
80 × 50
0.2
379.2 (±97.5)
122.1 (±44.21)
314.3 (±61.37)
111.0 (±57.14)
0.3
364.7 (±103.82)
125.7 (±54.5)
309.5 (±69.09)
121.5 (±43.5)
50 × 50
0.2
354.7 (±55.41)
130.5 (±25.5)
297.7 (±38.50)
122.4 (±20.7)
0.3
377.6 (±94.9)
147.2 (±35.06)
316.0 (±84.68)
129.1 (±28.5)
SD, standard deviation; FOV, field-of-view; MAR, metal artifact reduction.
Fig. 4SD of gray values (variability) with metal artifact reduction (MAR) and without MAR (w/o) under different conditions of voxel and field-of-view (FOV) sizes: A, Dental amalgam alloy, and B, Cu-Al alloy. Central bar = median; whiskers = first and third quartiles.
Assessment of bucal marginal alveolar peri-implant and periodontal defects using a cone beam CT system with and without the application of metal artefact reduction mode.
found that MAR reduced the diagnostic accuracy in the detection of vertical root fractures. The inconsistency between objective and subjective assessments on the efficacy of MAR highlights the importance of evaluating possible factors that may compromise the action of this tool. Considering that scanning parameters should be selected according to the diagnostic task, this study was conducted to evaluate the effect of the tool in different FOV and voxel sizes.
It is known that FOV size influences CBCT image quality due to the effect of projection data discontinuity,
Relationship between density variability and imaging volume size in cone-beam computerized tomographic scanning of the maxillofacial region: an in vitro study.
voxel size influences the detection of X-ray photons. Larger voxels produce images of lower spatial resolution, but can detect a greater number of X-ray photons, which results in higher signal and, consequently, less image noise.
Interestingly, the voxel sizes used in this study did not influence the performance of MAR.
When the dental materials were compared separately, less image noise was observed in the presence of Cu-Al alloy. This can be related to the physical characteristics of this material, considering that it is composed of chemical elements with low atomic number (Cu: 29 and Al: 13) in relation to dental amalgam (Ag: 47 and Hg: 80), which reduces artifact formation.
In this particular CBCT unit, it could be observed that MAR was applied after automatic thresholding of the original image based on the gray values of the artifacts, followed by image correction. This is also suggested by Bechara et al.,
which reinforces the idea that MAR is a post-processing tool. Considering that voxel size is related to image reconstruction rather than image acquisition in Picasso Trio, and this can be associated with FOV size, the authors considered it very pertinent to evaluate the influence of FOV and voxel sizes on MAR. Such information is important for a clinical situation, in which the practitioner can make use of MAR, bearing in mind that this will not negatively interact with the CBCT scanning protocol.
It is important to highlight that laboratory studies, as the present study, allow the researchers to have important control over the methodological conditions, so that the variables that can influence the object of study are controlled in vitro. This makes possible the isolate assessment of MAR on CBCT images of different FOV and voxel sizes without secondary interferences from the imaging phantom; however, such homogeneous interaction with the X-ray photons is not representative of a clinical condition.
In conclusion, the efficacy of MAR was similar in CBCT images of dental materials obtained with different FOV and voxel sizes. This reinforces the concept that imaging protocols should be selected on a case-by-case basis.
References
Scarfe W.C.
Farman A.G.
Sukovic P.
Clinical applications of cone-beam computed tomography in dental practice.
Detection of periimplant fenestration and dehiscence with the use of two scan modes and the smallest voxel sizes of a cone-beam computed tomography device.
Oral Surg Oral Med Oral Pathol Oral Radiol.2013; 115: 121-127
Assessment of bucal marginal alveolar peri-implant and periodontal defects using a cone beam CT system with and without the application of metal artefact reduction mode.
Relationship between density variability and imaging volume size in cone-beam computerized tomographic scanning of the maxillofacial region: an in vitro study.
The authors declare that they have no conflict of interest. The results presented here were presented, in part, at the 20th International Congress of Dentomaxillofacial Radiology, Santiago, Chile and 20th meeting of the Brazilian Association of Oral Radiology, Sao Paulo, Brazil.
30094-9&id=gr1.jpg){kind=link}
30094-9&id=gr2.jpg){kind=link}
30094-9&id=gr3.jpg){kind=link}
30094-9&id=gr4.jpg){kind=link}