A comparison of the diagnostic accuracy of cbct images of different voxel resolutions used to detect simulated small internal resorption cavities

  • 413 views
Uploaded on

 

  • Full Name Full Name Comment goes here.
    Are you sure you want to
    Your message goes here
    Be the first to comment
    Be the first to like this
No Downloads

Views

Total Views
413
On Slideshare
0
From Embeds
0
Number of Embeds
0

Actions

Shares
Downloads
8
Comments
0
Likes
0

Embeds 0

No embeds

Report content

Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

Cancel
    No notes for slide

Transcript

  • 1. doi:10.1111/j.1365-2591.2010.01749.x A comparison of the diagnostic accuracy of CBCT images of different voxel resolutions used to detect simulated small internal resorption cavities ˘ K. Kamburoglu & S. Kursun Department of Oral Diagnosis and Radiology, Faculty of Dentistry, Ankara University, Ankara, Turkey Abstract 5-point scale for detection of internal resorption cavities. Kappa coefficients were calculated to assess ˘ Kamburoglu K, Kursun S. A comparison of the diagnostic intra- and inter-observer agreement. Az values were accuracy of CBCT images of different voxel resolutions used to calculated and compared, with a significance level of detect simulated small internal resorption cavities. International a = 0.05. Endodontic Journal, 43, 798–807, 2010. Results Kappa values were fair-to-moderate for the Aim To assess ex vivo the use of two different cone- low-resolution Iluma images and good to excellent for beam computerized tomography (CBCT) units at differ- all other image sets. Both intra- and inter-observer ent voxel resolutions for the detection of simulated agreement values were higher for the Accuitomo small internal resorption cavities. images than for the Iluma images. The highest Az Methodology A total of 60 single-rooted mandib- values were obtained with the Accuitomo 0.125 mm3 ular incisors and canines were split mesiodistally images and the lowest with the Iluma 0.3 mm3 images. along the coronal plane into labial and lingual The Az values for the Accuitomo and Iluma ultra- and sections. Artificial internal root resorption was sim- high-resolution images were higher than those of the ulated by drilling with a 0.5-mm-diameter round bur low-resolution Iluma images. No differences (P > 0.05) to its full depth at either the cervical or apical for internal resorption detection were found between portion of the labial wall. In total, 60 cavities were the Az values of the Accuitomo 0.125 mm3, Accuito- created (30 cervical and 30 apical), and 60 locations mo 0.160 mm3, ultra-resolution Iluma and high-reso- (30 cervical and 30 apical) were left without cavities. lution Iluma images. The sections were rejoined and placed in dry human Conclusion Ultra-and high-resolution Iluma and mandibles in groups of 6 and imaged using two Accuitomo 0.125 and 0.160 mm3 CBCT images per- different CBCT units at different resolutions to obtain formed similarly and better than low-resolution Iluma five image sets as follows: Iluma, ultra-resolution images in the detection of simulated internal resorption (0.1 mm3); Iluma, high-resolution (0.2 mm3); Iluma, ex vivo. low-resolution (0.3 mm3); 3D Accuitomo 170, 6 · Keywords: cone-beam computerized tomography, 6 cm FOV (0.125 mm3); 3D Accuitomo 170, detection, internal resorption, voxel resolution. 8 · 8 cm FOV (0.160 mm3). Each image set was viewed by two observers and evaluated using a Received 13 February 2010; accepted 14 April 2010 Introduction Resorption of tooth structures consists of the dissolu- ˘ Correspondence: Kıvanc Kamburoglu, Oral Diagnosis and ¸ tion of mineralized tissues such as dentine–cementum Radiology Department, Faculty of Dentistry, Ankara Univer- sity, Ankara, Turkey (Tel.: +90 312 2965632; fax: +90 312 or alveolar bone by osteoclasts or osteoclast-type cells 2123954; e-mails: dtkivo@yahoo.com; kamburogluk@ (Kleoniki et al. 2002) because of physiological, patho- dentistry.ankara.edu.tr). logical, or idiopathic factors (Tronstad 1988, Hamilton798 International Endodontic Journal, 43, 798–807, 2010 ª 2010 International Endodontic Journal
  • 2. ˘ Kamburoglu & Kursun Detection of internal resorption by CBCT& Gutmann 1999). Root resorption is classified accord- another used a 3-D reconstructive method in the ex vivoing to its location as either internal or external assessment of an extracted tooth with internal resorp-resorption (Frank & Torabinejad 1998). With the tion (Kleoniki et al. 2002).exception of pulpectomized deciduous teeth, internal Cone-beam computerized tomography (CBCT) dedi-resorption is a rare pathological finding (Wedenberg & cated to dento-maxillofacial imaging was introduced inLindskog 1985) in which odontoblasts and predentine response to the high demand for a technique that couldare destroyed by multinucleated cells as a result of provide three-dimensional data at a lower cost andchronic pulp infection or trauma. In general, internal with lower absorbed doses than the conventional CTresorption is asymptomatic. Radiographically, it ap- used in medical radiology (Tyndall & Rathore 2008).pears as a well-demarcated, round-to-oval radiolucent The use of CBCT in clinical practice offers a number ofwidening or ballooning out of the pulpal canal that is potential advantages over conventional tomography,most frequently observed in the cervical region, but including easier image acquisition, higher image accu-which might occur in all areas of the root canal system racy, reduced artefacts, lower effective radiation doses,(Gunraj 1999, Ne et al. 1999, Patel & Dawood 2007). faster scan times and greater cost-effectiveness (ScarfeRadiographic detection of an internal resorptive defect et al. 2006, Scarfe & Farman 2008). Rather than theis important as it is an indication for pulpectomy. fan-shaped beam emitted by conventional CT technol-Without early diagnosis, prognosis is poor (Wedenberg ogy, CBCT units, as the name implies, emit a cone-& Lindskog 1985, Bakland 1992, Trope 2002). shaped X-ray beam. Because the beam covers the entire With the development of new diagnostic methods region of interest, images can be acquired in only oneand the increasing importance given to pulpal diseases, pass or less around the patient’s head. After the X-raythe diagnosis of internal resorption has become more beam passes through the patient, the remnant beam iscommon in routine dental practice (Patel et al. captured on a 2-D planar detector – either an amor-2009a,b). Conventional and digital intra-oral radiog- phous silicon flat panel or an image intensifier/CCDraphy are most commonly used for this purpose (White detector (White 2008) – and records a cylindrical or& Pharoah 2008). Orthoradial, mesioradial and disto- spherical volume of data in a field of view ranging fromradial views taken with a 20° variation in horizontal 4 to 30 cm. Voxels are isotropic and typically range inangle might assist in detecting and managing internal size between 0.08 and 0.4 mm3 (Patel et al. 2009a). ˘resorption (Kleoniki et al. 2002, Kamburoglu et al. Recent studies indicate that CBCT might be used2008a). With external resorption, the area of radiolu- successfully in the diagnosis of root resorption. CBCTcency will show movement when multiple radiographs was found to be more sensitive than panoramicare taken from mesial and distal angulations, whereas imaging in the detection of simulated canine-inducedwith internal resorption, the defect will continue to external root resorption (Alqerban et al. 2009). Aappear close to the canal (Frank & Torabinejad 1998, recent in vivo study also found CBCT to be superior toKleoniki et al. 2002). The detection of internal resorp- intra-oral radiography in the detection and manage-tion using conventional intra-oral techniques is also ment of external cervical and internal resorption (Patelaffected by the size and location of the lesion, with small et al. 2009b). Moreover, authors of a recent studyapical defects the most difficult to diagnose (Kambu- suggested a desktop cone-beam micro-CT scanner ˘roglu et al. 2008a). Regardless of the intra-oral system (Scanco Medical, Bassersdorf, Switzerland) to be aused, the two-dimensional nature of the images limits promising technology for measuring the diameter andthe information that can be obtained with regard to the volume of simulated internal resorptive defects (Kam-size, extent and location of internal resorption, and their ˘ buroglu et al. 2008b).diagnostic value is dependent upon beam angulation, Voxel size is of paramount importance in terms ofsuperimposition of anatomical structures and patient- quality and scanning and reconstruction times of CBCTrelated factors (Bjorndal et al. 1999, Patel & Dawood ¨ images. The influence of voxel size on the diagnostic ˘2007, Kamburoglu et al. 2008a). Three-dimensional ability of a CBCT unit (i-CAT; Imaging Sciencesvisualization has been proposed as a means of over- International, Hatfield, PA, USA) in evaluating simu-coming the limitations of intra-oral systems in the lated external root resorption has been assessed in aimaging of internal resorptive defects. For example, one previous laboratory study (Liedke et al. 2009); how-study suggested rotational tomography as an appropri- ever, no data exist on the effects of voxel size on theate method for recording a bucco-lingual view of a tooth detection of internal resorption. This study aimed atwith internal resorption (Friedland et al. 2001), and assessing the use of different voxel resolutions of twoª 2010 International Endodontic Journal International Endodontic Journal, 43, 798–807, 2010 799
  • 3. ˘ Detection of internal resorption by CBCT Kamburoglu & Kursun different CBCT units in the detection of internal volume of reconstruction up to 21.1 · 14.2 cm; and resorption ex vivo. a 3D Accuitomo 170 (3D Accuitomo; J Morita Mfg. Corp., Kyoto, Japan) with a flat-panel detector offering five different FOVs – 40 · 40, 60 · 60, 80 · 80, Material and methods 100 · 100 and 170 · 120 mm. With the Iluma sys- A total of 60 single-rooted mandibular incisors and tem, images were obtained at 120 kVp, 3.8 mA and an canines (20 central incisors, 20 lateral incisors, 20 exposure time of 40 seconds and were reconstructed at canines) with radiographically visible root canals and ultra- (0.1 mm3), high- (0.2 mm3) and low-resolution with no restorations, root fillings, pathosis or anomalies (0.3 mm3) voxel sizes. With the Accuitomo system, were split along the mesiodistal coronal plane into images were obtained at 65 kV, 2.0 mA and an labial and lingual sections using a low-speed saw exposure time of 30.8 seconds with a 60 · 60 mm (Isomet; Buehler Ltd., Evanston, IL, USA). Artificial FOV (0.125 mm3 voxel size) and at 65 kV, 2 mA and internal resorption was simulated by drilling with a an exposure time of 30.8 sec with an 80 · 80 mm FOV 0.5-mm-diameter round bur to its full depth under (0.160 mm3 voxel size). Axial scans and multiplanar ·2.5 magnification at either the cervical or apical reconstructions were obtained, and volumetric data portion of the labial wall. In total, 60 cavities were were reconstructed using the system software pro- created (30 cervical and 30 apical), and 60 locations grams to provide serial coronal and sagittal sections (30 cervical and 30 apical) were left without cavities. along each tooth plane. A total of 5 image sets were Teeth were obtained from individuals who gave obtained as follows: (i) Accuitomo, 6 · 6 cm FOV informed consent to donate their extracted teeth for (0.125 m3); (ii) Accuitomo, 8 · 8 cm FOV (0.160 research purposes. mm3); (iii) Iluma, ultra-resolution (0.1 mm3); (iv) Tooth segments were rejoined using Super Glue Iluma, high-resolution (0.2 mm3); (v) Iluma, low- (Scotch, Super Glue gel; 3M, St Paul, MN, USA) and resolution (0.3 mm3) (Figs 1–5). placed in the appropriate alveolar sockets of two dry Each image set was evaluated separately in a human mandibles in groups of 6 (two incisors, two random order by two calibrated observers using the laterals, two canines), making a total of 10 groups. software and enhancement tools of the imaging Images were obtained from two different CBCT units: systems. Image sets were viewed at 1-week intervals, an Iluma Ultra Cone-beam CT Scanner (3M Imtec, and evaluations of each image set were repeated Ardmore, OK, USA) with a 24.4 · 19.5 cm amorphous 1 month after the initial viewings. The presence or silicon flat-panel image detector and a cylindrical absence of internal resorption (a demarcated widening Figure 1 Coronal and cross-sectional images of a canine tooth with apical internal resorption cavities shown by the arrows reconstructed from Accuitomo, 6 · 6 cm FOV (0.125 m3).800 International Endodontic Journal, 43, 798–807, 2010 ª 2010 International Endodontic Journal
  • 4. ˘ Kamburoglu & Kursun Detection of internal resorption by CBCT (a) (b) (c)Figure 2 (a) Axial, (b) coronal, (c) cross-sectional view of a tooth reconstructed from Accuitomo, 8 · 8 cm FOV (0.160 mm3).Arrows show cervical internal resorption cavities in coronal and cross-sectional images.Figure 3 Serial cross-sectional images of an incisor tooth with apical internal resorption cavities reconstructed by Iluma, ultra-resolution (0.1 mm3). Apical internal resorption cavities are visible on some sections.of the pulpal canal) was scored using the following 5- a screen resolution of 1440 · 900 pixels and 32-bitpoint scale: 1 = lesion definitely present; 2 = lesion colour depth.probably present; 3 = uncertain/unable to tell; 4 = le- Weighted kappa coefficients were calculated to assesssion probably not present; 5 = lesion definitely not intra- and inter-observer agreement for each image set.present. No time restriction was placed on the To take into account the clustered structure of the dataobservers. Images were viewed in a dimly lit room resulting from the assessment of two regions (apicalon a 22¢¢ LG Flatron monitor (LG, Seoul, Korea) set at and cervical) for each tooth, Clustered Data Statisticalª 2010 International Endodontic Journal International Endodontic Journal, 43, 798–807, 2010 801
  • 5. ˘ Detection of internal resorption by CBCT Kamburoglu & Kursun Figure 4 Cross-sectional images of the same tooth shown in Fig. 3 reconstructed from Iluma, high-resolution (0.2 mm3). Apical internal resorption cavities are visible on some sections. Figure 5 Cross-sectional images of the same tooth shown in Figs 3 and 4 reconstructed from Iluma, low-resolution (0.3 mm3). Please note that apical internal resorption cavities are not visible.802 International Endodontic Journal, 43, 798–807, 2010 ª 2010 International Endodontic Journal
  • 6. ˘ Kamburoglu & Kursun Detection of internal resorption by CBCTSoftware (Ankara University, Department of Biostatis- images for either observer. Statistically significanttics, Ankara, Turkey) was used to assess the accuracy differences were found between Az values for theof each image system/voxel size by calculating and Accuitomo 0.125 mm3 and low-resolution Ilumacomparing the areas under the receiver operating images for both observer 1 (P = 0.0004) and observercharacteristic (ROC) curves (Az values) for each image 2 (P = 0.0005). Statistically significant differencesset/observer, with the significance level set at a = 0.05. were also found between the Az values of the low- resolution (0.3 mm3) Iluma images and those of the Accuitomo 0.160 mm3 (P = 0.0019), high-resolutionResults Iluma (0.2 mm3) (P = 0.02) and ultra-resolution Ilu-Intra-observer and inter-observer kappa coefficients ma (0.1 mm3) (P = 0.0084) images for observer 2.calculated by image type are shown in Tables 1 and Table 4 shows the sensitivity, specificity, positive pre-2, respectively. Intra-and inter-observer kappa values dictive value (PPV) and negative predictive value (NPV)for both sets of Accuitomo images and for the ultra- calculated for both observers. Figures 6 and 7 show theand high-resolution Iluma images were good, whereas ROC curves drawn for observer 1 and observer 2,kappa values for the low-resolution Iluma images were respectively.only fair-to-moderate. Both intra- and inter-observeragreement values were higher for the Accuitomo Discussionimages than for the Iluma images. Given the overall acceptable intra-observer agree- This study found both ultra-and high-resolution Ilumament, Az values were calculated for each image set and and 0.125 mm3 and 0.160 mm3 Accuitomo CBCTobserver using the first readings only (Table 3). For images were better than low-resolution Iluma CBCTboth observers, the Accuitomo 0.125 mm3 images images for the detection of artificially created internalwere found to have the highest Az values, whereas the root resorption. No statistically significant differencesIluma low-resolution Iluma images were found to have were found between the ultra- and high-resolutionthe lowest Az values. No statistically significant differ- Iluma and Accuitomo images. Both intra- and inter-ences (P > 0.05) were found amongst the Az values of observer agreement values were relatively better withthe Accuitomo 0.125 mm3, Accuitomo 0.160 mm3, the Accuitomo images. The highest Az values, sensi-Iluma ultra-resolution and Iluma high-resolution tivity and kappa values were obtained with the 0.125 mm3 (6 · 6 cm FOV) Accuitomo images. This study evaluated the effects of voxel size on theTable 1 Intra-observer kappa coefficients calculated by detection of internal root resorption using two differentimage type CBCT systems. Small-sized cavities were used, as Weighted kappa-SE previous studies have shown them to be the most Observer 1 Observer 2 difficult to detect using intra-oral radiographs (Holmes ˘ et al. 2001, Kamburoglu et al. 2008a). A previousIluma 0.3 mm3 0.465–0.091 0.524–0.090 laboratory study (Liedke et al. 2009) assessing theIluma 0.2 mm3 0.734–0.087 0.772–0.091Iluma 0.1 mm3 0.807–0.090 0.878–0.090 influence of voxel size on the diagnostic ability of aAccuitomo0.160 mm3 0.854–0.091 0.915–0.091 CBCT unit (i-CAT; Imaging Sciences International) inAccuitomo0.125 mm3 0.878–0.091 0.917–0.091 evaluating simulated external root resorption found that voxel resolution did not affect sensitivity orTable 2 Inter-observer kappa coefficients calculated by specificity values. However, 0.3 mm3 voxel resolutionimage type was associated with greater diagnostic performance and lower patient exposure when compared to 0.2 and Weighted kappa-SE 0.4 mm3 voxel resolutions (Liedke et al. 2009). In 1st readings 2nd readings contrast, the present study found the low-resolution Obs 1–Obs 2 Obs 1–Obs 2 (0.3 mm3) Iluma images to be the least effective of allIluma 0.3 mm 3 0.364–0.069 0.457–0.080 systems tested in terms of accuracy and observerIluma 0.2 mm3 0.561–0.091 0.565–0.091 agreement. The difference in results between the twoIluma 0.1 mm3 0.650–0.090 0.637–0.091 studies might be because of the use of larger cavitiesAccuitomo 0.160 mm3 0.766–0.091 0.837–0.091 (0.6, 1.2 and 1.8 mm diameter) in the earlier study,Accuitomo 0.125 mm3 0.886–0.091 0.854–0.091 which would be easier to detect at all voxel resolutions,ª 2010 International Endodontic Journal International Endodontic Journal, 43, 798–807, 2010 803
  • 7. ˘ Detection of internal resorption by CBCT Kamburoglu & Kursun Table 3 Az values for the first readings of both observers, by image type Observer 1 Observer 2 Image type Az value SE 95% CI Az value SE 95% CI 3 Iluma 0.3 mm 0.6908 0.0414 (0.6074–0.7743) 0.6542 0.0435 (0.5666–0.7417) Iluma 0.2 mm3 0.7711 0.0359 (0.6989–0.8433) 0.7812 0.0393 (0.7021–0.8604) Iluma 0.1 mm3 0.7861 0.0364 (0.7128–0.8594) 0.8115 0.0409 (0.7292–0.8939) Accuitomo 0.160 mm3 0.7994 0.0398 (0.7193–0.8796) 0.8353 0.039 (0.7567–0.9138) Accuitomo 0.125 mm3 0.8686 0.0285 (0.8112–0.926) 0.8492 0.0357 (0.7772–0.9211) Table 4 Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) for both observers Sensitivity Specificity PPV NPV Image type Obs 1 Obs 2 Obs 1 Obs 2 Obs 1 Obs 2 Obs 1 Obs 2 3 Iluma 0.3 mm 0.57 0.63 0.82 0.62 0.76 0.63 0.66 0.63 Iluma 0.2 mm3 0.65 0.70 0.85 0.75 0.81 0.74 0.71 0.72 Iluma 0.1 mm3 0.62 0.73 0.93 0.85 0.90 0.83 0.71 0.76 Accuitomo 0.160 mm3 0.67 0.75 0.90 0.88 0.87 0.86 0.73 0.78 Accuitomo 0.125 mm3 0.87 0.75 0.85 0.90 0.85 0.89 0.87 0.82 ROC Curve ROC Curve 1.0 1.0 0.8 0.8 Sensitivity 0.6 0.6 Sensitivity Source of the Source of the 0.4 curve 0.4 curve Iluma 0.3 Iluma 0.3 Iluma 0.2 Iluma 0.2 Iluma 0.1 Iluma 0.1 0.2 Accuitomo 0.160 0.2 Accuitomo 0.160 Accuitomo 0.125 Accuitomo 0.125 Reference line Reference line 0.0 0.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Specificity Specificity Figure 6 Receiver operating characteristic (ROC) curves for Figure 7 Receiver operating characteristic (ROC) curves for observer 1 for each image type. observer 2 for each image type. and to the fact that masking by the root canal has less with the Accuitomo images. Moreover, observers noted of an effect on the detection of external root resorption that the Accuitomo system had more user friendly than on the detection of internal root resorption. In software when detecting small internal resorption addition, the present study found that the use of smaller cavities in terms of image reconstruction and viewing FOVs resulted in higher Az values. The fact that the capabilities. In this study, different image types were Iluma system does not offer multiple FOVs is a evaluated in different sessions by using dedicated significant drawback that might explain the relatively software of both systems in a random order. Totally, lower Az values obtained with the Iluma in comparison randomizing the images would have been impractical804 International Endodontic Journal, 43, 798–807, 2010 ª 2010 International Endodontic Journal
  • 8. ˘ Kamburoglu & Kursun Detection of internal resorption by CBCTand cause recall of the images of the same tooth leading Effective doses from the Accuitomo CBCT have beento observer bias. reported to vary between 11 and 77 lSv, depending on In clinical practice, a variety of additional factors FOV, exposure parameters and region examined. Inaffect the detection of root resorption using CBCT, comparison with other regions of the jaw, imaging ofincluding observer performance, viewing conditions, the central incisor region results in lower effectiveCBCT hardware and software specifications, beam- doses, making it possible to image this region usinghardening artefacts and patient-related factors. Obser- lower exposure parameters than those required forver ability in this study might have been enhanced by other regions (Lofthag-Hansen et al. 2008). Thethe use of artificial defects, which have clear borders Accuitomo 170 system used in this study offers fivethat are easier to detect than those of natural defects. different FOVs; however, only the 6 · 6 cmMoreover, patient motion, which represents a signifi- (0.125 mm3 voxel resolution) and 8 · 8 cm FOVscant drawback in the clinical use of CBCT, was not a (0.160 mm3 voxel resolution) were assessed. Thefactor in this laboratory study. Similarly, image degra- results of the present study suggest that not only dodation and beam hardening and scatter related to soft smaller FOVs offer the advantage of lower effectivetissue, filling material and restorations in the imaged doses, they also enable better detection. It is possiblearea were not an issue. that the 4 · 4 FOV (0.08 mm3 voxel resolution), the The use of CBCT images in the in vivo detection and smallest offered by the Accuitomo system, would havemanagement of root resorption lesions was compared enhanced the diagnostic ability of the observers;with intra-oral periapical radiography in an earlier however, because it was not possible to image all thestudy conducted with 15 patients (five with internal 6 teeth simultaneously using the 4 · 4 FOV, it was notresorption, five with external cervical resorption and assessed in this study.five without resorption) (Patel et al. 2009b). The unit Whilst the Iluma unit tested does not offer multipleassessed (3D Accuitomo 80; J Morita) was an earlier FOVs, it is possible to reconstruct images at 0.1, 0.2model of the Accuitomo CBCT used in this study. Intra- and 0.3 mm3 after exposure. According to ICRP 2007oral radiography yielded Az values of 0.780 and 0.830 tissue weights, Iluma images acquired at 120 kVp andfor the diagnostic accuracy of internal and external 3.8 mA (as in our study) result in an effective dose ofcervical resorption, respectively, whereas Az values 498 lSv (Ludlow & Ivanovic 2008). This is higherwith CBCT images were 1.0 for both internal and than the effective doses from full-mouth radiographsexternal cervical resorptions. Sensitivity, specificity, (FMX) taken with photostimulable phosphor (PSP)PPV and NPV were also 1.0 with CBCT, and both storage and F-speed film with rectangular collimationintra-and inter-observer agreement values were higher (34.9 lSv), FMX taken with PSP and F-speed film withfor CBCT images than for intra-oral radiography. The round collimation (170.7 lSv), and FMX taken withauthors concluded that CBCT images were superior to D-speed film and round collimation (388 lSv) (Ludlowintra-oral radiography in the diagnosis and manage- et al. 2008). With regard to the system’s 0.1 mm3ment of both internal and external resorption. In voxel resolution, in addition to the relatively highcomparison with the earlier study, this study yielded effective dose, the long reconstruction times associatedlower Az values, sensitivity, specificity, PPV and NPV. with this resolution precludes its use in routine clinicalThis finding is likely because of the small size of the dental imaging. It is apparent that Accuitomo outper-simulated lesions assessed in this study. In spite of this, forms Iluma in terms of effective doses.with the exception of the low-resolution (0.3 mm3) ˘ Another study (Kamburoglu et al. 2008b) found aIluma images, the Az values, sensitivity, specificity, desktop cone-beam micro-CT (Scanco Medical, Bassers-PPV, NPV and intra-and inter-observer kappa scores dorf, Switzerland) with a 16-lm (0.016 mm3) isotropicobtained in this study can be considered highly resolution to be a promising technology for measuringacceptable. the diameter and volume of simulated internal resorp- This study was not designed to measure the effective tive defects in extracted tooth. However, the ultra-highdoses derived from the different FOVs on the equipment radiation dose, high cost and long scanning and imagetested. However, a recent study found smaller FOVs reconstruction times associated with micro-CT make itresulted in lower effective radiation doses and suggested ˘lu unsuitable for clinical use (Kamburog et al. 2008b).that as a rule, smaller FOVs should be used for dental Future research should be directed towards developingimaging, with the use of larger FOVs restricted to cases units that offer higher-definition dental images withwhere a wider view is required (Hirsch et al. 2008). lower effective doses.ª 2010 International Endodontic Journal International Endodontic Journal, 43, 798–807, 2010 805
  • 9. ˘ Detection of internal resorption by CBCT Kamburoglu & Kursun Conclusion cavities – an ex vivo study in human cadaver jaws. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology and In the ex vivo detection of simulated internal root Endodontology 105, 790–7. resorption, the highest Az values were obtained with ˘ Kamburoglu K, Barenboim SF, Arıturk T, Kaffe I (2008b) ¨ Accuitomo 0.125 mm3 images and the lowest with Quantitative measurements obtained by micro-computed Iluma 0.3 mm3 images. Ultra-and high-resolution tomography and confocal laser scanning microscopy. Iluma and Accuitomo 0.125 and 0.160 mm3 CBCT Dentomaxillofacial Radiology 37, 385–91. images performed similarly and better than low-resolu- Kleoniki ML, Vasiliki ID, Ourania CHP, Theodoros L, Ioannis tion Iluma CBCT images. Both intra- and inter-observer KP (2002) Internal root resorption studied by radiography, agreement values were relatively higher for the Accui- stereomicroscope, scanning electron microscope and com- puterized 3D reconstructive method. Dental Traumatology tomo images when compared to the Iluma images. 18, 148–52. Liedke GS, Dias da Silveira HE, Dias da Silveira HL, Dutra V, de Acknowledgements Figueiredo JA (2009) Influence of voxel size in the diagnostic ability of cone beam tomography to evaluate The authors thank Bilal Arkan for technical assistance simulated external root resorption. Journal of Endodontics 35, and Deborah Semel for editorial assistance. 233–5. Lofthag-Hansen S, Thilander-Klang A, Ekestubbe A, Helmrot E, Grondahl K (2008) Calculating effective dose on a cone ¨ References beam computed tomography device: 3D Accuitomo and 3D Alqerban A, Jacobs R, Souza PC, Willems G (2009) In-vitro Accuitomo FPD. Dentomaxillofacial Radiology 37, 72–9. comparison of 2 cone-beam computed tomography systems Ludlow JB, Ivanovic M (2008) Comparative dosimetry of and panoramic imaging for detecting simulated canine dental CBCT devices and 64-slice CT for oral and maxillo- impaction-induced external root resorption in maxillary facial radiology. Oral Surgery, Oral Medicine, Oral Pathology, lateral incisors. American Journal of Orthodontics and Dento- Oral Radiology and Endodontology 106, 106–14. facial Orthopedics 136, 764.e1–11. Ludlow JB, Davies-Ludlow LE, White SC (2008) Patient risk Bakland LF (1992) Root resorption. Dental Clinics of North related to common dental radiographic examinations: the America 36, 491–507. impact of 2007 International Commission on Radiological Bjorndal L, Carlsen O, Thuesen G, Darvann T, Kreiborg S ¨ Protection recommendations regarding dose calculation. (1999) External and internal macromorphology in 3D- Journal of American Dental Association 139, 1237–43. reconstructed maxillary molars using computerized X-ray Ne RF, Witherspoon DE, Gutmann JL (1999) Tooth resorption. microtomography. International Endodontic Journal 32, 3–9. Quintessence International 30, 9–25. Frank AL, Torabinejad M (1998) Diagnosis and treatment of Patel S, Dawood A (2007) The use of cone beam computed extracanal invasive resorption. Journal of Endodontics 24, tomography in the management of external cervical resorp- 500–4. tion lesions. International Endodontic Journal 40, 730–7. Friedland B, Faiella RA, Bianchi J (2001) Use of rotational Patel S, Dawood A, Whaites E, Pitt Ford T (2009a) New tomography for assessing internal resorption. Journal of dimensions in endodontic imaging: part 1. Conventional Endodontics 27, 797–9. and alternative radiographic systems. International Endodon- Gunraj MN (1999) Dental root resorption. Oral Surgery, Oral tic Journal 42, 447–62. Medicine, Oral Pathology, Oral Radiology and Endodontology Patel S, Dawood A, Wilson R, Horner K, Mannocci F (2009b) 88, 647–53. The detection and management of root resorption lesions Hamilton RS, Gutmann JL (1999) Endodontic-orthodontic using introral radiography and cone beam computed relationships: a review of integrated treatment planning tomography – an in vivo investigation. International End- challenges. International Endodontic Journal 32, 343–60. odontic Journal 42, 831–8. Hirsch E, Wolf U, Heinicke F, Silva MA (2008) Dosimetry of Scarfe WC, Farman AG (2008) What is cone-beam CT and how the cone beam computed tomography Veraviewepocs 3D does it work? Dental Clinics of North America 52, 707–30. compared with the 3D Accuitomo in different field of views. Scarfe WC, Farman AG, Sukovic P (2006) Clinical applications Dentomaxillofacial Radiology 37, 268–73. of cone-beam computed tomography in dental practice. Holmes JP, Gulabivala K, van der Stelt PF (2001) Detection of Journal of Canadian Dental Association 72, 75–80. simulated internal tooth resorption using conventional Tronstad L (1988) Root resorption – etiology, terminology and radiography and subtraction imaging. Dentomaxillofacial clinical manifestations. Dental Traumatology 4, 241–52. Radiology 30, 249–54. Trope M (2002) Root resorption due to dental trauma. ˘ Kamburoglu K, Barenboim SF, Kaffe I (2008a) Comparison of Endodontic Topics 1, 79–100. conventional film with different digital and digitally filtered Tyndall DA, Rathore S (2008) Cone-beam CT diagnostic images in the detection of simulated internal resorption applications: caries, periodontal bone assessment, and806 International Endodontic Journal, 43, 798–807, 2010 ª 2010 International Endodontic Journal
  • 10. ˘ Kamburoglu & Kursun Detection of internal resorption by CBCT endodontic applications. Dental Clinics of North America 52, White SC (2008) Cone-beam imaging in dentistry. Health 825–41. Physics 95, 628–37.Wedenberg C, Lindskog S (1985) Experimental internal White SC, Pharoah MJ (2008) The evolution and application resorption in monkey teeth. Endodontics and Dental Trauma- of dental maxillofacial imaging modalities. Dental Clinics of tology 1, 221–7. North America 52, 689–705.ª 2010 International Endodontic Journal International Endodontic Journal, 43, 798–807, 2010 807