doi:10.1111/iej.12390

CBCT-based volume of simulated root resorption – influence of FOV and voxel size

P. F. Da Silveira, M. P. Fontana, H. W. Oliveira, M. B. Vizzotto, F. Montagner, H. L. Silveira & H. E. Silveira Department of Surgery and Orthopedics, Oral Radiology Division, Federal University of Rio Grande do Sul, Porto Alegre, Brazil

Abstract Da Silveira PF, Fontana MP, Oliveira HW, Vizzotto MB, Montagner F, Silveira HL, Silveira HE. CBCTbased volume of simulated root resorption – influence of FOV and voxel size. International Endodontic Journal.

Aim To examine the influence of the field of view (FOV) and voxel size on the measurement of the volume of simulated internal root resorption (IRR) lesions through cone-beam computed tomography (CBCT). Methodology Eleven single-rooted teeth with IRR simulated by acid demineralization were studied. CBCT images were acquired using large FOV (voxel sizes of 0.200, 0.250 and 0.300 mm) and limited FOV (voxel sizes of 0.076, 0.100 and 0.200 mm). The IRR volumes were calculated using the Dolphinâ software. Volumetric measurements were validated using IRR silicone putty casts. The analysis of variance (ANOVA) for randomized block design complemented with the Tukey’s test was employed. Results IRR volumes obtained using voxel sizes of 0.200 and 0.250 mm were similar (P > 0.05). How-

Introduction Internal root resorption (IRR) is a lesion associated with an inflammatory process of the pulp. The prognosis of IRR is complex, and its diagnosis is based on

Correspondence: Priscila Fernanda Da Silveira, Dental School – Federal University of Rio Grande do Sul, Oral Radiology Division, Rua Ramiro Barcelos, 2492, Porto Alegre, RS, CEP 90035-003, Brazil (Tel/Fax: 55 51 3308 5199; e-mail: [email protected]).

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

ever, both these values were significantly different from that obtained using the 0.300-mm voxel (P < 0.05). There was no significant difference between IRR volumes measured through voxel sizes of 0.076 and 0.100 mm (P > 0.05), but both differed significantly from that obtained through the 0.200mm voxel (P < 0.05). There was no significant difference between the volumetric measurements of the 0.200-mm voxel images of the restricted and large FOV protocols. The mean volumes of the silicone casts were smaller than those calculated using a 0.200mm voxel, but were similar to those obtained using voxel sizes of 0.076 and 0.300 mm. Conclusions Despite the FOV protocol, voxel size can influence measurement of simulated IRR volumes. The importance of standardization of CBCT image acquisition protocols is emphasized, especially during follow-up of an IRR lesion, to prevent misinterpretation of its extent, which can create a bias in clinical decisions. Keywords: cone-beam computed measurements, root resorption.

tomography,

Received 14 April 2014; accepted 30 September 2014

clinical and imaging examinations. Early detection of IRR is important for achieving a good treatment outcome (Fuss et al. 2003). Several studies have demonstrated the accuracy of cone-beam computed tomography (CBCT) in detecting the early stages of IRR when compared to other imaging techniques such as conventional radiographs (Kamburoglu & Kursun 2010, Kamburoglu et al. 2011). Apart from the improved diagnostic ability, CBCT also allows three-dimensional analysis of IRR lesions. It can provide further data on the extent of damage that IRR can cause to teeth. Recent studies have

International Endodontic Journal

1

CBCT volume of root resorption Da Silveira et al.

employed CBCT-based volumetric measurements to evaluate different diseases as well as anatomic structures such as the maxillary sinus and airways (Ahlowalia et al. 2013, Esposito et al. 2013, Ghoneima & Kula 2013, Kim et al. 2013, Stoetzer et al. 2013). The analysis is facilitated by specific software-based tools that can calculate volumes from CBCT data. CBCT measurements follow a 1 : 1 scale and are accurate in the assessment of both length and thickness (Sherrard et al. 2010). The accuracy of volume tools in software such as Dolphin 3D (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA) has been demonstrated in some investigations (Sherrard et al. 2010, Ye et al. 2012, Ghoneima & Kula 2013). However, these studies utilized high-resolution images. The influence of voxel size on image resolution has been widely confirmed in the literature, especially in relation to diagnosis of different diseases (Spin-Neto et al. 2012). Small voxel sizes generate images with high diagnostic power that can detect several changes such as root fractures, canals and root resorption (Liedke et al. 2009, Neves et al. 2012, Da Silveira et al. 2013, Ponder et al. 2013, Vizzotto et al. 2013). Differences have been observed in the quality of images obtained from different CBCT devices when the field of view (FOV) or the voxel size is changed (Spin-Neto et al. 2012). However, the influence of image acquisition parameters on volumetric measurements has not been established yet. Volumetric measurements achieved through CBCT can be additional tools for evaluating the extent of IRR. It is important to standardize imaging protocols to determine the real volume of the lesion and consequently allow proper treatment planning for each specific clinical situation. There is a lack of literature comparing different CBCT-based acquisition protocols and their relationship with volumetric analysis tools in the diagnosis of IRR. The present study investigated the influence of FOV and voxel size on CBCTbased volumetric measurement of simulated IRR. Therefore, the null hypothesis considered that there was no difference amongst the IRR volumes when obtained through different CBCT protocols.

Materials and methods

2

single-rooted human teeth were selected for the IRR simulation protocol, as described by Da Silveira et al. (2014). Briefly, all teeth were sectioned using a diamond disk (Buehler Diamond Cut-Off Wheels 114243; Buehler, Lake Bluff, IL, USA) coupled to an electric saw (Isomet Low Speed Saw; Buehler). Root extension was measured using a digital caliper (Mitutoyo Sul Americana Ltda, Santo Amaro, SP, Brazil). After determining the middle third of the root, IRR was simulated using 5% nitric acid and 8% sodium hypochlorite. After 11 days, cavities of 11 different sizes that simulated IRR were obtained.

Image acquisition Tomographic images of all the teeth in the sample were acquired. The teeth were placed in alveoli that were constructed using a dried mandible covered with a wax layer to simulate soft tissues. The same positions of the teeth were maintained throughout the process of image acquisition. Images were acquired using 2 CBCT devices with three voxel size protocols (Table 1). Therefore, each tooth was scanned using six different CBCT protocols.

Volume calculations The Digital Imaging and Communications in Medicine (DICOM) format files were exported to the Dolphin 3D software (Dolphin Imaging and Management Solutions, Chatsworth, CA, USA) and visualized on a 22inch flat-screen monitor (Flatron E2250, 1920 9 1080 dpi; LG, Taubate, SP, Brazil). A trained and calibrated examiner conducted the IRR volumetric measurements. These measurements were repeated at an interval of at least 1 week, and the intraclass correlation coefficient (ICC) was calculated as 0.999 (P < 0.001). The sinus/airway tool was used to define and calculate the volumes. The sensitivity of the tool was carefully determined for each type of image, and the operator manually defined the extent of each lesion, performing adjustments in the sagittal, coronal and axial slices, as shown in Fig. 1. Thereafter, volumetric measurements of the IRR simulations were automatically determined in mm³ by the software.

Sample preparation

Validation of volumetric measurements

This study was approved by the Ethics Committee in Research, Federal University of Rio Grande do Sul (UFRGS), Porto Alegre, RS, Brazil. Eleven extracted

To validate volumetric measurements calculated using the Dolphin 3D software and correlate them with a clinical situation, the IRR simulations were modelled

International Endodontic Journal

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

Da Silveira et al. CBCT volume of root resorption

Table 1 CBCT devices and the voxel size, FOV, current, voltage, and scanning time protocols used in this study Voxel (mm)

Group i-Cat Next Generation (Imaging Sciences International, Inc., Hatfield, PA, USA) Kodak 9000 3D (KODAK Dental Systems, Carestream Health, Rochester, NY, USA)

(a)

(b)

(c)

(d)

0.200 0.250 0.300 0.076 0.100 0.200

FOV (cm)

Voltage (kV)

Current (mA)

Scan time (s)

6 9 16

120

5 9 3.7

60

37.07 37.07 18.54 8

26.9 26.9 8.9 18.8

(e)

Figure 1 CBCT images (i-Cat, 0.200 mm) of one tooth of the sample analysed using the Dolphin 3D software tool for calculation of the simulated IRR volume; (a) sagittal section, IRR visualization; (b) axial section, IRR delimitation and volume filling by the software; (c) sagittal section; (d) coronal section; (e) volumetric reconstruction of the tooth using the IRR volume.

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

International Endodontic Journal

3

CBCT volume of root resorption Da Silveira et al.

and subsequently, volumes were calculated. Six additional single-rooted teeth were used for this purpose. Middle- and large-sized IRR lesions were simulated separately in two samples of three teeth each, as described in a previous study (Da Silveira et al. 2014). Cavities were moulded using light body silicone impression putty (Elite HD+; Zhermack SpA, Badia Polesine, Italy). The low-viscosity silicone putty was carefully injected directly inside the cavities using an automix syringe. At the end of the filling procedure, the segmented teeth were repositioned. On completion of the curing time of the impression, putty casts of the cavities were obtained. The simulated cavity casts were weighed in precision scales (Schimadzu Corporation, Kyoto, Japan), and thereafter, volumes were calculated by dividing the weight of the cavity cast by the density of the impression putty (d = 1.40 g cc 1), as described in previous studies (Ahlowalia et al. 2013, Esposito et al. 2013, Ghoneima & Kula 2013). The values thus obtained were converted to mm³ to allow comparisons to be made with the values obtained using CBCTbased volumetric measurements. CBCT images of the two samples were then acquired using the same protocols. Thereafter, cavity volumes were calculated using the volume calculation tool and compared to those obtained using the silicone putty casts. The comparison of values thus obtained is presented in Table 3.

Statistical analysis The IRR volumetric measurements of the teeth obtained using the 2 CBCT devices and different acquisition protocols were analysed statistically. The analysis of variance (ANOVA) for randomized block design complemented with the Tukey’s test was employed. The significance level was set at 5%.

detected between the mean IRR simulation volumes obtained using voxel sizes of 0.076 mm (9.25 mm³) and 0.100 mm (9.58 mm³); however, these values were significantly different from those obtained using the 0.200-mm voxel size protocol (10.73 mm³). The IRR volumes obtained with the 0.200-mm voxel size using i-Cat Next Generation and Kodak 9000 3D devices were not significantly different (P < 0.05). The mean volumes of the IRR cavities modelled with silicone putty were similar to the mean volumes obtained using the 0.076 and 0.300-mm voxel CBCT devices. The results also revealed that the mean value of the IRR obtained with the 0.200-mm voxel size from both CBCT units was 14.23 mm3. The mean value obtained from the silicone putty casts was 10.31 mm3. The results are shown in Table 3.

Discussion The difference between various imaging techniques and the diagnostic ability of a specific method can influence the identification of an IRR. The determination of the IRR landscape and size may influence the treatment strategy. Volume reconstruction tools with CBCT imaging have been extensively used (Ahlowalia et al. 2013, Esposito et al. 2013, Ghoneima & Kula 2013, Kim et al. 2013, Stoetzer et al. 2013) to enhance the visualization of other clinical conditions. The calculation of the lesion’s volume is frequently considered a complementary analysis in the imagebased diagnosis through multiplanar reconstructions. The efficiency of different CBCT-based protocols may directly affect the diagnostic ability of the examination and the quality of the image (Spin-Neto et al. Table 2 Mean and standard deviation of the volumetric measurement of IRR simulations obtained from images acquired using different CBCT devices and voxel sizes Volume (mm³)

Results Group

Table 2 presents the mean volumes of the IRR simulations in the 11 teeth measured using 2 CBCT devices with different FOV and voxel sizes. The mean cavity volumes measured using the i-Cat Next Generation device did not differ significantly between the 0.200 (11.31 mm³) and 0.250 (11.38 mm³) voxel sizes; however, these values differed significantly (P < 0.03) from the values obtained using a 0.300mm voxel size (7.95 mm³). When the Kodak 9000 3D device was employed, no significant difference was

4

International Endodontic Journal

i-Cat Next Generation

Kodak 9000 3D

Voxel 0.200 0.250 0.300 0.076 0.100 0.200

Mean Aa

11.31 11.38A 7.95B 9.25B 9.58B 10.73Aa

SD 1.89 1.86 1.67 1.58 1.62 1.75

Means followed by different uppercase letters for the same CBCT device and means followed by same lowercase letters for all CBCT devices indicate statistically significant differences in ANOVA with randomized block design complemented with the Tukey’s test (significance level: 5%).

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

Da Silveira et al. CBCT volume of root resorption

Table 3 Volume (mm³) of the IRR simulation in teeth, measured using silicone putty casts and CBCT images of different voxel sizes, for the validation of volumetric measurements Volume (mm³) i-Cat Next Generation

Tooth 01 02 03 04 05 06 Mean

Weight (g) 0.0200 0.0214 0.0165 0.0129 0.0094 0.0064

g g g g g g

Density (g cc 1) 1.40

Kodak 9000 3D

Cast

0.200-mm voxel

0.250-mm voxel

0.300-mm voxel

0.076-mm voxel

0.100-mm voxel

0.200-mm voxel

14.29 15.29 11.77 9.21 6.71 4.57 10.31

18.3 19.3 16.5 13.7 9.9 7.7 14.23

17.2 19.2 15 12.9 10.1 6.4 13.47

13.1 16 11.9 10 6.6 3.7 10.22

12.9 15 9.8 9 6.3 3.7 9.45

17.9 19.8 12.9 12.2 8.7 6.1 12.9

18.1 18.9 15 11 8.9 6.8 13.2

2012). However, the interference of these variables in CBCT-based imaging for the volumetric determination of IRR has not been assessed. The present study aimed to evaluate the influence of CBCT acquisition parameters in IRR volume calculations using a specific image tool. Images of simulated IRR lesions were acquired with different voxel size protocols using two different CBCT devices: the i-Cat Next Generation (a large FOV CBCT device) and the Kodak 9000 3D (a restrict FOV device). No difference was observed in the IRR volumes produced by the i-Cat Next Generation CBCT when the images were obtained with the 0.200- and 0.250-mm voxel sizes. Both protocols produced images of comparable quality for volumetric measurements, corroborating earlier studies that have reported similar results for volumetric measurements and diagnostic ability of these voxel sizes, suggesting that these protocols can produce similar results (Liedke et al. 2009, Ye et al. 1012, Spin-Neto et al. 2012, Da Silveira et al. 2013, Vizzotto et al. 2013). This could be explained by the similar voltage, amperage and acquisition time applied in both these protocols. However, the 0.300-mm voxel images of IRR had smaller volumes than the 0.200 and 0.250 mm protocols. Current literature reports that the voxel size can influence the resolution of CBCT images and consequently, their diagnostic ability (Liedke et al. 2009, Neves et al. 2012, Spin-Neto et al. 2012, Da Silveira et al. 2013, Vizzotto et al. 2013). The 0.300 mm protocol produced small images because a lower number of images were required to generate the IRR image when compared to other protocols. Low-resolution CBCT images have more noise and dissimilar grey scale shades (Katsumata et al. 2007,

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

Schulze et al. 2011, Azeredo et al. 2013). In a previous study, the volumes of roots were measured using CBCT images acquired using different voxel sizes, and the authors observed that a larger voxel size resulted in larger tooth volume that was not representative of the real root size (Ye et al. 2012). This could be related to the poor image definition seen in protocols using larger voxel sizes. Ponder et al. (2013) measured external apical root resorption defects through CBCT images and concluded that a 0.200 mm voxel size provided more accurate measurements than the 0.400 mm one. In the present study, the simulated IRR tended to be smaller in the 0.300-mm voxel images than in those of a 0.200-mm voxel size. The possible explanation is the lower definition on the edges of the simulated IRR lesions observed in images of a larger voxel size (0.300 mm). Images obtained using the Kodak 9000 3D tomography device for both 0.076- and 0.100-mm voxel sizes showed no significant difference in volumes. However, the mean values for the 0.200-mm voxel size images were higher than the previous ones. Although all these images were obtained using highresolution protocols, current literature shows that reducing voxel size below 0.200 mm increases image noise than impairs the observation of low-contrast areas, especially smaller ones (Tanimoto & Arai 2009, Costa et al. 2012). In the present study, using either restrict or large FOV did not influence the IRR volume measured using 0.200-mm voxel images. Therefore, it may be suggested that image quality is more closely associated with the voxel size than the FOV size. For the same voxel protocol, restrict FOV seems to be adequate, especially for small areas, because these

International Endodontic Journal

5

CBCT volume of root resorption Da Silveira et al.

protocols were obtained with a low radiation exposure dose (Scarfe et al. 2012). According to the ‘as low as is reasonably achievable’ (ALARA) principle, an adequate diagnosis should be obtained by exposing patients to the lowest possible radiation dose. The validation of volumetric measurements through CBCT images and silicone putty casts as a measure of real volume have been reported by other authors (Ahlowalia et al. 2013, Esposito et al. 2013, Ghoneima & Kula 2013). However, these studies used only 1 CBCT acquisition protocol, and, in some cases, the voxel size was not described. In this study, it was observed that volumetric measurements of the silicone putty casts were similar to those obtained using the 0.076 and 0.300 mm resolution voxel. The 0.200mm voxel images of both CBCT units were larger than the silicone putty casts. The boundaries of an IRR simulated by the acid demineralization technique present a dentine layer with low mineral content, and therefore a small density when viewed in the tomographic image, which could be interpreted as a void by the software but not by the silicone putty casts. The validation model that employed silicone casts may not be the best method to calculate the volumes of lesions produced by acid demineralization. Silicone casts are useful for well-defined structures or cavities produced with drills, but do not capture the natural architecture of IRR inside the root canal. Moreover, the IRR lesions produced by an acid protocol may be closer to those produced by the in vivo process. There is a need to employ effective imaging protocols that provide detailed information on the real volume of IRR lesions. The present study demonstrated that low-resolution and small voxel size protocols can underestimate the real dimensions of the IRR. The precise characterization of IRR lesions allows determining their relation to the dentine and periodontal tissues. These findings are important because they will guarantee the use of proper clinical treatment measures and avoid misdiagnosis. Adopting different acquisition protocols in CBCT might generate volume differences, resulting in under- or overestimation of the IRR lesion size. This information is important not only for determining the initial diagnosis but also for the follow-up period.

Conclusion The use of protocols with different voxel sizes in CBCT may significantly change volumetric measurements in IRR simulated lesions, whereas there appears to be

6

International Endodontic Journal

little influence of FOV on the IRR volumetric determination. Voxel size should be standardized in CBCT image acquisition protocols to avoid under- or overestimation of the lesion size, which could influence the decision-making process in the clinical treatment of IRR.

Acknowledgements The authors deny any conflict of interests. We affirm that we have no financial affiliation (e.g. employment, direct payment, stock holdings, retainers, consultantships, patent licensing arrangements or honoraria), or involvement with any commercial organization with direct financial interest in the subject or materials discussed in this manuscript, nor have any such arrangements existed in the past three years. Any other potential conflict of interest is disclosed.

References Ahlowalia MS, Patel S, Anwar HMS et al. (2013) Accuracy of CBCT for volumetric measurement of simulated periapical lesions. International Endodontic Journal 46, 538–46. Azeredo F, Menezes LM, Enciso R, Weissheimer A, Oliveira RB (2013) Computed gray levels in multislice and conebeam computed tomography. American Journal of Orthodontics and Dentofacial Orthopedics 144, 147–55. Costa FF, Gaia BF, Umetsubo OS, Pinheiro LR, Tortamano IP, Cavalcanti MGP (2012) Use of Large-volume ConeBeam Computed Tomography in Identification and Localization of Horizontal Root Fracture in the Presence and Absence of Intracanal Metallic Post. Journal of Endodontics 38, 856–9. Da Silveira PF, Vizzotto MB, Liedke GS, da Silveira HL, Montagner F, da Silveira HE (2013) Detection of vertical root fractures by conventional radiographic examination and cone beam computed tomography - an in vitro analysis. Dental Traumatology 29, 41–6. Da Silveira PF, Vizzotto MB, Montagner F, Silveira HLD, Silveira HED (2014) Development of a new in vitro methodology to simulate internal root resorption. Journal of Endodontics 2, 211–6. Esposito SA, Huybrechts B, Slagmolen P et al. (2013) A novel method to estimate the volume of bone defects using cone-beam computed tomography: an in vitro study. Journal of Endodontics 9, 1111–5. Fuss Z, Tsesis I, Lin S (2003) Root resorption–diagnosis, classification and treatment choices based on stimulation factors. Dental Traumatology 19, 175–82. Ghoneima A, Kula K (2013) Accuracy and reliability of cone-beam computed tomography for airway volume analysis. European Journal of Orthodontics 35, 256–61.

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

Da Silveira et al. CBCT volume of root resorption

Kamburoglu K, Kursun S (2010) A comparison of the diagnostic accuracy of CBCT images of different voxel resolutions used to detect simulated small internal resorption cavities. International Endodontic Journal 43, 798–807. Kamburoglu K, Kursun S, Yuksel S, Oztas B (2011) Observer ability to detect ex vivo simulated internal or external cervical root resorption. Journal of Endodontics 37, 168–75. Katsumata A, Hirukawa A, Okumura S et al. (2007) Effects of image artifacts on gray-value density in limited-volume cone-beam computerized tomography. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and Endodontics 104, 829–36. Kim ES, Moon SY, Kim SG, Park HC, Oh JS (2013) ThreeDimensional Volumetric Analysis After Sinus Grafts. Implant Dentistry 22, 170–4. Liedke GS, da Silveira HE, da Silveira HL, Dutra V, de Figueiredo JA (2009) Influence of voxel size in the diagnostic ability of cone beam tomography to evaluate simulated external root resorption. Journal of Endodontics 35, 233–5. Neves FS, Freitas DQ, Campos PSF, de Almeida SM, HaiterNeto F (2012) In vitro comparison of cone beam computed tomography with different voxel sizes for detection of simulated external root resorption. Journal of Oral Science 54, 219–25. Ponder SN, Benavides E, Kapila S, Hatchd NE (2013) Quantification of external root resorption by low- vs high-resolution cone-beam computed tomography and periapical radiography: a volumetric and linear analysis. American Journal of Orthodontics and Dentofacial Orthopedics 143, 77–91.

© 2014 International Endodontic Journal. Published by John Wiley & Sons Ltd

Scarfe WC, Li Z, Aboelmaaty W, Scott SA, Farman AG (2012) Maxillofacial cone beam computed tomography: essence, elements and steps to interpretation. Australian Dental Journal 57(1 Suppl), 46–60. Schulze R, Heil U, Gross D et al. (2011) Artifacts in CBCT: a review. Dento Maxillo Facial Radiology 40, 265–73. Sherrard JF, Rossouw E, Benson BW, Carrillo R, Buschange PH (2010) Accuracy and reliability of tooth and root lengths measured on cone-beam computed tomographs. American Journal of Orthodontics and Dentofacial Orthopedics 137, S100–8. Spin-Neto R, Gotfredsen E, Wenzel A (2012) Impact of voxel size variation on CBCT-based diagnostic outcome in dentistry: a systematic review. Journal of Digital Imaging 4, 813–20. Stoetzer M, Nickel F, Rana M et al. (2013) Advances in assessing the volume of odontogenic cysts and tumors in the mandible: a retrospective clinical trial. Head & Face Medicine 9, 14. Tanimoto H, Arai Y (2009) The effect of voxel size on image reconstruction in cone-beam computed tomography. Oral Radiology 25, 149–53. Vizzotto MB, Silveira PF, Ar us NA, Montagner F, Gomes BPFA, Silveira HED (2013) CBCT for the assessment of second mesiobuccal (MB2) canals in maxillary molar teeth: effect of voxel size and presence of root filling. International Endodontic Journal 9, 870–6. Ye N, Jian F, Xue J et al. (2012) Accuracy of in-vitro tooth volumetric measurements from cone-beam computed tomography. American Journal of Orthodontics and Dentofacial Orthopedics 142, 879–87.

International Endodontic Journal

7

CBCT-based volume of simulated root resorption - influence of FOV and voxel size.

To examine the influence of the field of view (FOV) and voxel size on the measurement of the volume of simulated internal root resorption (IRR) lesion...
237KB Sizes 2 Downloads 7 Views