This study assessed the effect of amperage (mA) and voltage (kVp) alterations in CBCT with a small FOV on metal artifact generation in the exomass, and the CBCT image quality. The present study is among the first to assess the effect of artifacts generated in the exomass area on CBCT image quality. Although metal artifacts in CBCT have been the topic of many investigations, studies on quantitative assessment of the effect of such artifacts on image quality are scarce [18]. Qualitative assessments by the examiners may be used to assess the amount of artifacts and their diagnostic impacts; however, such assessments cannot be used as a basis for the purpose of comparison of the efficacy of different protocols, and are not suitable for quality control either. There is no standard parameter for quantitative assessment of the effect of metal objects [19]. Metal artifacts have been quantified by using the mean and SD of VG in the literature. The mean GV provides an overall estimate of the rate of darkness and lightness of an image due to the presence of metals. In areas with low GV due to beam hardening artifact, an increase in the mean GV and a reduction in its SD would indicate a reduction in metal artifacts. Higher SD values indicate higher noise and lower image quality [20]. Thus, the mean and SD of GV were calculated in the present study.
The present results showed that voltage, amperage, and type of metal object significantly affected the amount of artifacts. Presence of metal objects in the exomass area and lower amperage decreased image quality. However, unilateral/bilateral presence of metal objects in the exomass area alone had no significant effect on the mean GV and image quality; although its interaction with the type of metal object had a significant effect on the GV and image quality.
Precise selection of exposure parameters, such as voltage and amperage plays an important role in enhancement of image quality and reduction of patient radiation dose. It has been reported that CBCT with high exposure parameters decreases the noise and improves image homogeneity, which results in more accurate image reconstruction of tissues with close density values, and enables the observer to differentiate between close GVs. On the other hand, high exposure parameters increase the patient radiation dose [21]. Nonetheless, high voltage and amperage improve image quality. Thus, exposure parameters should be optimized to acquire the highest image quality with the lowest patient radiation dose possible. Oliveira et al. [12] reported that changing the amperage and voltage affected the artifacts generated in the exomass; however, the effect of changing the voltage was greater than the amperage on exomass artifacts. Their results were in agreement with the present findings. It should be noted that increasing the voltage decreases the mean GV, which is in contrast to the abovementioned statement but the obtained image is more homogenous than when a lower voltage is used. In other words, by increasing the voltage, image contrast decreases and the GVs of gray shadows are approximated. Resultantly, the mean GV is lower compared with when lower voltage is used. Another advantage of increasing the voltage in CBCT scanners is the reduction in patient radiation dose [21]. Thus, increasing the voltage yields a more homogenous image with less artifacts and lower patient radiation dose.
In the present study, increasing the amperage significantly decreased the exomass artifacts but its effect was smaller than the effect of voltage. Increasing the amperage increases the mean GV and yields a more homogenous image. However, it also increases the patient radiation dose which is not favorable. The present results also revealed the significant interaction effect of amperage and voltage, such that the lowest mean GV was obtained in use of 90 kVp and 10 mA; in other words, increasing the voltage and amperage improved the image quality and decreased artifacts.
The present study also assessed the interaction effect of confounding factors on the GV and image quality. The results revealed that simultaneous alterations of voltage and amperage had the greatest effect on image quality; the interaction effect of type of metal object and voltage ranked next. Moreover, the results revealed the poorest image quality in presence of amalgam in the exomass area, compared with the other two metal objects. Safi et al. [22] showed that type and number of metal objects affected the artifacts caused by the metal objects in the exomass area.
The present study indicated that type of metal object affected the artifacts generated in the exomass area; however, number of metal objects had no significant effect on artifact generation. The reason may be the alterations in exposure parameters and using the smallest size of FOV because changing the amperage and voltage affects the image quality and GV. Candemil et al. [15] concluded that presence of metal objects in the endomass or their simultaneous presence in the endomass and exomass areas significantly decreased the mean GV and increased the noise, which was in agreement with the present findings. Kocasarac et al. [23] reported that implants present in the exomass yielded images with higher SD and artifacts than implants located within the FOV, which was in line with the present findings. Katsumata et al. [24] evaluated the relationship of GV and CBCT volume size using Alphard CBCT scanner, which can provide different imaging volume sizes. They demonstrated the significant effect of data discontinuity in CBCT scans with limited volume size. Although increasing the image volume size had the reliable advantage of higher density, it increased the patient radiation dose. Also, larger imaging volume, due to larger voxel size, decreases image resolution and results in loss of details [24]. Moreover, previous studies demonstrated that increasing the size of FOV decreased the variability in GVs [24–26]. In the present study, a 5 x 5 cm FOV, which is among the smallest FOVs of CBCT scanners was used; however, the results indicated that reduction of voltage and amperage decreased the image quality. Addition of metal objects to the exomass also decreased the image quality. Thus, it may be concluded that in case of presence of a limitation in selection of the size of FOV, better quality images may be obtained by paying attention to the exposure parameters and also type and number of metal objects in the exomass area. Oliveira et al. [12] found that the variability of voxel values decreased in presence of exomass, which was in contrast to the present findings. The exomass was a homogenous and thin layer of water in their study, which might have served as an X-ray filter and increased the mean energy of X-ray beams, resulting in lower level of changes within the FOV. Such findings highlight the significance of exomass in designing methods for assessment of CBCT GV [12].
Considering the common use of metallic materials with high density and atomic number in dentistry, presence of such materials in the exomass would decrease the image quality [16, 27]. Presence of metal objects in the exomass creates some hypodense and hyperdense streaks that are in fact beam hardening artifacts [19, 20]. In the present study, presence of metal objects in the exomass decreased the GV, and amalgam build-up restoration caused the maximum reduction in GV. This finding indicates that the hypodense artifacts are dominant and in a clinical situation, they probably compromise correct detection of hypodense structures such as a fracture line. Nonetheless, a previous study reported that presence of high-density materials in the exomass and their negative impact on image quality did not affect the diagnostic accuracy for detection of artificially induced vertical root fractures [28]. This difference in the results was probably due to the different type of scanner, and constant optimized exposure parameters in their study. Candemil et al. [16] showed that presence of any metal object in the exomass in use of CS9300 and PicassoTrio scanners and presence of two Co-Cr and amalgam cylinders in use of NewTom Giano CBCT scanner decreased the mean GV, which was in agreement with the present findings. However, in contrast to the present findings, they showed that presence of one to three Co-Cr or amalgam cylinders in scanning with NewTom Giano caused hyperdense artifacts. This difference in the results may be due to different exposure parameters (amperage and voltage) of different CBCT scanners. Also, it has been discussed that the artifacts generated by objects with different linear attenuation coefficients are not the same in use of different CBCT scanners [29].
In the present study, the lowest mean GV and the highest SD of GV were recorded for amalgam, followed by Co-Cr post, and titanium implant. This result can be attributed to different atomic numbers of these metal objects. The atomic number of titanium implant is 22, the atomic number of Cr and Co is 24 and 27, respectively, and that of mercury, silver, tin, and copper is 80, 47, 50, and 29, respectively [11, 16]. Thus, higher atomic number and physical density increase artifact generation following X-ray exposure [5, 30], which has a direct correlation with lower mean and higher SD of GV. Increased heterogeneity of the GV indicates greater effect of artifacts on CBCT images and reduction of image quality [19, 28].
In the present study, the mean GV of titanium implants in bilateral position in different voltage and amperage values was lower than that of Co-Cr post. Although Co-Cr post has a higher atomic number than titanium implant, it should be noted that the diameter of titanium implants used in the present study was higher than that of Co-Cr posts to better simulate the clinical setting. Thus, irrespective of the lower density of titanium implant, its larger diameter probably increases the artifacts following X-ray exposure [15, 22].
Candemil et al. [11] demonstrated that without using the metal artifact reduction algorithm in PicassoTrio and ProMax scanners, presence of Co-Cr alloy in the exomass area caused further reduction in the mean GV, and further increase in SD of GV, compared with titanium, which was consistent with the present findings. Another study evaluated the effect of presence of three titanium implants in the exomass using Cranex, 3Dx, Orthopos SL-3D, and XI scanners and reported a significant reduction in the mean GV compared with the control position [27]. Consistent with the present findings, Safi et al. [22] reported the highest reduction in the mean GV due to the presence of amalgam alloy, and the lowest reduction due to the presence of titanium implants in the exomass area.
In addition to the type of metal object, number of metal objects was also evaluated in the present study, and the results showed no significant difference in reduction of GV in unilateral and bilateral presence of metal objects. However, previous studies reported that higher number of metal objects in the exomass decreased the mean GV and increased the SD of GV [22, 24, 25, 31], which was in contrast to the present findings. It should be noted that previous studies assessed the effect of number of metal objects on the mean and SD of GV in different sizes of FOV. Evidence shows that by a reduction in size of FOV, the effects of metal artifacts in the exomass increase, causing a reduction in the mean and an increase in SD of GV. However, one size of FOV was used in the present study to assess the pure effect of amperage and voltage on the number and type of metal objects. It appears that the effect of size of FOV on the number of metal objects in the exomass is greater than the effect of other exposure parameters; however, this topic was not assessed in the present study. Thus, it may be stated that in case of presence of higher number of metal objects in the exomass, selection of a larger FOV probably has a greater effect on reduction of artifacts due to metal objects in the exomass compared with changing the amperage and voltage.
A 5 x 5 cm FOV, which is one of the smallest FOVs of CBCT scanners was used in the present study, which was a strength. Size of FOV is one of the most important factors affecting the patient radiation dose [16, 24]. Using the smallest possible FOV has been recommended to protect against ionizing radiation and improve image quality [30, 32]. However, in CBCT, reduction in size of FOV in the axial plane indirectly increases the exomass effect and occurrence of truncation [22, 24, 31, 33], and heterogeneity in the reconstructed volume [34]. It should be noted that all previous studies on exomass artifacts used metal objects of the same volume and dimensions [11, 15, 16, 27, 28, 35, 36], which is different from the clinical setting. The present study used teeth with amalgam build-up and Co-Cr intracanal posts and titanium implants to better simulate the clinical setting, which was another strength of the current study. The majority of available studies on artifacts generated by metal objects in the exomass used polypropylene cylinder phantoms [4, 11, 35]; while, metal objects were mounted in a dry human mandible in the present study to better simulate the clinical setting [15]. Furthermore, ballistic gelatin was used for soft tissue simulation, which has been reported to be the best material for this purpose [17]. Nonetheless, due to in vitro design, generalization of results to the clinical setting should be done with caution as the result of inter-individual differences [37]. Also, motion artifact, which is a common problem in the clinical setting, was not simulated in this in vitro study. Future studies are recommended on other high-density artifact-generating materials using different CBCT scanners with different image reconstruction algorithms. Moreover, evaluation of quantitative parameters related to the amount of artifacts along with qualitative assessment by experienced experts can increase the practicality of the results of quantitative studies. Finally, the present results should be verified in clinical studies.