Accuracy of Cone-Beam Computed Tomography in Detection of Root Perforations Using Different Voxel Sizes in Comparison to Digital Periapical Radiograph, An in vitro study

doi:10.21203/rs.3.rs-3318405/v1

Objectives:

The aim of this study is to quantitatively compare the diagnostic accuracy of CBCT with 0.2 and 0.4 mm³voxel sizes and a digital PR for the detection of strip perforation for mesial root of the mandibular first molar with and without root canal obturation when evaluated by different observer specialty.

Materials and Methods:

48 intact mandibular first molars were selected for this study, the teeth were divided to the following groups: (i) no strip perforation and no obturation, (ii) no strip perforation and obturation, (iii) strip perforation and no obturation, (iv) strip perforation and obturation. Each tooth was inserted in a dry mandible to obtain intraoral digital, CBCT images were obtained using two voxel sizes 0.2 and 0.4 mm³. Two general practitioners, two endodontists, and two dentomaxillofacial radiologists were asked to evaluate all images for detection of strip root perforation individually and together. The sensitivity and specificity of all images modalities to diagnose perforations were calculated.

Results:

There was no statistically significant difference between the accuracy of the PR images observed by the two general practitioners and two radiologists, while they showed a statistically low significant difference compared to the images observed by the two endodontists. CBCT images accuracy showed a higher significant difference than PR images observed by general practitioners and radiologists.

Conclusions:

CBCT was associated with significantly greater accuracy for detection of strip root perforation than PR. When two endodontist specialists observed the PR together, they showed a higher significance difference in the detection of root perforation compared to general practitioners and radiologists, and they showed no significant difference compared CBCT images. There was no significant difference between 0.2 and 0.4 mm³voxel sizes of CBCT images.

Clinical Relevance:

The accuracy of PR in the detection of root perforation is equal to that of CBCT when observed by two endodontists, and it is higher than the accuracy of PR images observed by general practitioners and radiologists.

cone-beam computed tomography

strip root perforation

diagnostic accuracy

periapical radiography

sensitivity

specificity

Endodontic treatment depends on many factors, one of them is the expertise of the clinician and the presence of root perforation. Mandibular first molars have the highest frequency of endodontic treatments [1]; the mesial root frequently has a curvature that makes endodontic treatment more challenging and increases the chance of procedure errors; root perforation could occur during root canal treatment in the thinner wall of the danger zone as a result of this anatomical curvature. Several factors could affect the prognosis of root perforation, including the time of diagnosis and the size of the perforation [2]; early determination of this complication is essential in terms of both deciding appropriate treatment procedures and preventing medico-legal actions [3]. Several instruments and techniques were employed to diagnose the clinical root perforations. These include a dental operative microscope [4], an endoscope [5], an electronic apex locator [6, 7], and an optical coherence tomography scan [8]. However, none of these could detect perforations in roots that were already filled because they relied on seeing the empty root canal space. In these situations, radiographic methods could be used; two-dimensional radiographs like panoramic and periapical provide 2D images of structures that may interfere with the detection of perforation because they superimpose intraoral anatomical structures; three-dimensional (3D) imaging like cone-beam computed tomography is the best method for the detection of perforations [9]. The quality of CBCT pictures is influenced by a number of variables, including voxel size, field of view (FOV), tube current (mA), and tube voltage (kaph). Voxels are the smallest volumetric elements in 3D images, and they have a substantial impact on the quality of scanning and the time of image acquisition [10, 11]. The as low as reasonably achievable (ALARA) rule states that the patient should receive the lowest radiation dosage possible to obtain appropriate image quality. Using a smaller voxel size increases the radiation dose and improve the image quality [12]. Therefore, in order to get the best images with the lowest radiation dose, balance should be achieved in accordance with the ALARA guideline [13]. The purpose of this study is to quantitatively compare the diagnostic accuracy of CBCT with different voxel sizes for the detection of root perforation in comparison to a digital periapical radiograph (PR) when evaluated by different observer specialty .

Aim of the work:

The aim of this study is to quantitatively compare the diagnostic accuracy of CBCT with 0.2 and 0.4 mm³voxel sizes and a digital PR for the detection of strip perforation for mesial root of the mandibular first molar with and without root canal obturation when evaluated by different observer specialty.

This study protocol was approved by the Institutional ethical board (no. KFSIRB200-1). In this study, it is necessary to include 48 intact mandibular first molars for 4 groups (n = 12) to be able to analyse inter-observer and intra-observer compatibility. Sound teeth that were extracted for periodontal reasons were selected and stored in distilled water at room temperature. Teeth with external or internal root resorption, anomalies, fractures, cracks, and immature apices were not included in the study. Any hard deposits on the teeth were removed using sharp curettes, and soft deposits were removed by soaking the teeth in a 5.25% sodium hypochlorite solution (NaOCl) for 30 minutes. The endodontic access cavities were prepared, and a #10 K-file was used to verify the canal's patency. The working length of the mesial roots was measured to be 1 mm short of the estimated true canal length after a #10 K-file was inserted into the canal until its tip was visible at the apical foramen. The two mesial root canals were instrumented in each tooth using Protaper Next rotary systems (PTN, Dentsply Maillefer, Ballaigues, Switzerland) up to X2 files (size 25, apical taper 6%). At each instrument change during the instrumentation procedure, the canals were irrigated with a 2.5% NaOCl solution. Following instrumentation; to make the canal efficiently clean and remove any remaining debris passive ultrasonic irrigation using an Irrisafe 20/21 file (Satelec, Merinac Cedex, France) was used. After that, 2 mL of distilled water was used to rinse the canals. Throughout the experiment, all teeth were kept moist with water.

The teeth were divided randomly into four groups (n = 12) according to presence or absence of strip perforation and presence or absence of root obturation as the follows: (i) no strip perforation and no obturation of the mesial canals, (ii) no strip perforation and obturation of the mesial canals, (iii) strip perforation and no obturation of the mesial canals, (iv) strip perforation and obturation of the mesial canals. To perform root canal perforation, Gates Glidden drill No. 1 (Dentsply Maillefer, Ballaigues, Switzerland) was used in the coronal third of the mesiolingual root canal until the perforation was made at the distal surface of the mesial canal. The perforation defect was confirmed by a size 20 K-file through the perforation area. For root canal obturation, root canals were dried using paper points. AH plus sealer was mixed according to the manufacturer’s instructions, and mater cone gutta-percha cone size 25 with a 0.04 taper was coated with the sealer and placed into the root canal. The roots were filled using the lateral condensation technique with a spreader and standard size 20 gutta-percha cones. After that, a hot instrument was used to cut off the gutta-percha to the level of the canal orifices. For radiographic techniques and evaluation, the teeth were inserted into prepared sockets in dry human mandibles in the posterior area. Three coats of dental wax were applied to the buccal and lingual surfaces of the mandibles to resemble soft tissue. The wax layer was 1.5 mm thick all around. To secure the teeth in the mandibular holes and fill spaces between the root surface and the socket, impression material (Elite HD, Zhermack, Italy) was utilised. Intraoral digital PR were taken using a Gendex X-ray unit (Gendex Digital Systems, Hatfield, PA) operating at 65 kVp, 7 mA, and 0.2 s exposure time. The GXPS-500 PSP (photostimulable phosphor plate) was used with a parallel technique and three horizontal angles, including the direct angle, 10^omesially, and 10^o distally (Figure 1). The visibility of enamel, dentine, and pulpal root canals were used as indicators for the best image quality. Additionally, using the ProMax® 3D Max CBCT equipment (Planmeca, Helsinki, Finland), CBCT images at two different voxel sizes 0.2 mm³ (CBCT-0.2) (Figure 2) and 0.4 mm³ (CBCT-0.4) (Figure 3) were obtained at 96 kVp, 1 mA, with a 55 x 50 mm FOV and exposure periods ranging between 12 and 15 s. The CBCT pictures were examined in the three reconstruction planes, comprising the axial, coronal, sagittal, and multiplanar reconstruction images with a slice thickness of 0.1 mm and an interval of 0.1 mm. A specific calibration session using 10 images that were not included in the study was conducted prior to the interpretation process. Image sets were viewed separately by six blinded, calibrated observers in intraoral digital PR and CBCT assessment, three pairs of observers participated in this study: two general practitioners, two endodontists, and two dentomaxillofacial radiologists. The observers did not participate in the samples preparation, and they were blind to the results of the other imaging technique. All images were randomised within each imaging technique; no time limit was placed on the observers. Image sets were viewed individually at 1-week intervals, and evaluations of each image set were repeated 1 week after the initial viewings, and if there was a difference between them, a final decision was asked for each observer independently, they examined the images using a monitor (LG Flatron 18.5-inch, Seoul, Korea) in a low-lit room and graded their observations as ‘perforation’ or ‘no perforation’. After that, each pair of observers discussed the cases together with the possibility of perforation until a consensus was reached.

Statistical analysis:

The data were collected and analysed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). Cohen’s kappa values (k) were used to calculate intra-observer (difference in repeated measurements by the same observer) and inter-observer agreement (difference in the measurements between two observers). Kappa values were interpreted as the following criteria: <0.10 = no agreement; 0.10–0.40 = poor agreement; 0.41–0.60 = moderate agreement; 0.61–0.80 = strong agreement; and 0.81–1.00 = excellent agreement [14]. The area under the receiver operating characteristic curves (AUC) with standard errorE) and 95% confidence interval (CI) was calculated; the AUC values under 0.5 correspond to scores with no discrimination ability, whereas values 1 correspond to perfect discrimination. AUC values for each image mode and observer were compared using z tests, and the significance level was set at p = 0.05. Se, Sp, NPV, and PPV with a confidence interval (CI 95%) of both CBCT scans and PR for the detection of the perforations were also calculated.

Table 1 shows the intra-observer kappa values. The values for the general practitioner observers while observing the digital PR were poor (k = 0.167 to 0.333), and they were moderate to excellent (k = 0.5 to 1) with CBCT-0.2 and CBCT-0.4, except for the 1^st general practitioner observer with images of CBCT-0.4 with non-obturated teeth was poor (k = 0.351). The endodontist observers showed excellent values of between-session consistency in all image evaluations (k = 0.83 to 1) except with the 1^st endodontist while observing the digital PR of non-obturated teeth was poor (k = 0.33). For radiologist observers, during observing the digital PR; the kappa values were moderate for non-obturated and obturated teeth (k = 0.437 and 0.5 respectively) with the 1^st radiologist observer and strong for non-obturated and obturated teeth (k = 0.625 and 0.8 respectively) with the 2^nd observer. While observing the CBCT-0.2, the value was moderate (k = 0.5) with the 1^st and 2^nd observers while observing the obturated teeth, and it was strong (k = 0.667) and excellent (k = 0.833) for the 1^st and 2^nd observers respectively with non-obturated teeth. The images for CBCT-0.4 showed poor agreement (k = 0.167 for non-obturated teeth and k = 0.333 for obturated teeth) with the 1^st observer, while it was strong (k = 0.667) and excellent (k = 1) with the 2^nd observer during observing the obturated and non-obturated teeth respectively. The interobserver agreements (Table 2) in the detection of perforations between the individual observers (1^st and 2^nd observers), were excellent for both CBCT voxel sizes (k = 0.833 to 1). For digital radiographic images, the kappa values were poor (k = 0.167 to 0.4) for individual readings with non-obturated teeth, for obturated teeth the value was poor between the two radiologists (k = 0.111), strong between the two general practitioners (k = 0.667), and excellent between the two endodontists reading (k = 1). Tables 3 and 4 show the AUC values, Se, Sp, PPV, and NPV for all observers and image sets for non-obturated and obturated teeth respectively. When the images were observed individually for non-obturated teeth, the calculated AUC values ranged between 0.583 and 0.833 for the digital PR images, and it was 1 for both CBCT voxel sizes images except for images observed by the 1^st general practitioner observer (AUC = 0.917). Statistical analysis for PR images of non-obturated teeth observed by the 1^st general practitioner showed a statistically low significant difference with the images observed by 2^nd observer (p < 0.001) and with the images observed by the pairs of observers together (p < 0.001), while for obturated teeth the AUC value was equal for individual observers and for the pairs of observes (AUC = 0.833). The AUC values for the PR images of non-obturated teeth observed by the 1^st and 2^nd endodontists showed the same values (AUC = 0.833) when they observed the images individually, and the results showed a statistically low significant difference when compared to the images observed by both endodontists together (p < 0.001), for the images of obturated teeth, the 1^st and 2^nd endodontist observers showed the same AUC values when they observed the images individually or together (AUC = 1). When the radiologists observed the digital PR images individually, the 1^st and 2^nd observers did not show significant differences from each other (p = 0.21 and 0.99 for non-obturated and obturated teeth respectively); however, both of their readings showed a statistically low significant difference compared to the observation of the two radiologists together with both non-obturated and obturated teeth (p < 0.001). When the digital PR images were observed by the pairs of observers together, the AUC values were 0.833 and 0.917 for the images observed by general practitioners and radiologists respectively, with no statistically significant difference between them (p = 0.084), the images observed by the endodontists (AUC = 1) showed higher significant difference compared to digital PR images observed by the general practitioners and radiologists for non-obturated and obturated teeth (p < 0.001). Only when the pairs of endodontists observe the images together; there was no statistically significant difference between PR and CBCT images (AUC = 1). The digital PR images showed sensitivity and specificity of 83.3% for general practitioner observers, and they showed 100% sensitivity and 83.3% specificity when observed by the radiologists for non-obturated and obturated teeth. The sensitivity and specificity for all CBCT images were 100% except for CBCT-0.4 images for non-obturated teeth observed by the 1^st general practitioner (sensitivity 83.3%) and CBCT-0.2 images for obturated teeth when observed by the 2^nd radiologist (specificity 83.3%). Comparing the CBCT images with the two different voxel sizes (0.2 and 0.4 mm³) did not show statistical significance with all observers except for the 1^st general practitioner with non-obturated teeth and the 2^nd radiologist with obturated teeth (p < 0.001).

In the current study, all observers investigated the diagnostic ability of CBCT taken in two different voxel sizes using a small fixed field-of-view in comparison to digital intraoral PR with a parallel technique and three horizontal angles for the detection of strip perforation for the mesial root of mandibular first molars with and without root canal obturation. In terms of decision-making and treatment planning, choosing the right imaging parameters is important for the precise identification of endodontic problems [15]. Mandibular first molars were chosen for the current study because of their greater morphological variety and more complex internal anatomy [16, 17], according to Berutti [18] the thinnest dentin thickness is situated 1.5 mm below the furcation area on the distal surface of the mesial root, and this area has a very high rate of perforation, which compromises the chances of a successful root canal and endangers the periradicular tissues health. In order to replicate the clinical condition in terms of the existence of cancellous and cortical bone, these molars were mounted on a human mandible [16], dental wax was applied on the mandibles buccally and lingually to replicate soft tissue in in vitro tests [19]. The most often used imaging modalities in endodontic clinical practice are PR [20, 21]. There has been a lot of debate on these two-dimensional images' limitations, which cause geometric distortion and restrict information related to the extension, size, and location of defects, various strategies have been suggested to enhance the abilities of conventional radiography [22]. Among the several imaging modalities, CBCT was created specifically to produce accurate 3D images of the maxillofacial skeleton, including the teeth and tissues that surround them [23]. According to the previous studies CBCT has the best accuracy for detecting perforations [16, 24]. The results of our study showed that CBCT images had higher inter-observer agreement values than 2D intraoral PR images and proved that CBCT images had great sensitivity and specificity in detecting root perforations, this might be due to the fact that superimposition of the neighbouring structures decreases the PR’s sensitivity, while the 3D nature of the CBCT enables visualisation of the perforation in different sections and angulations [25]. This findings are consistent with those of earlier researches [26, 27]. In our study, there were large inter- and intra-observer discrepancies during the interpretation of PR images by the general practitioners; this may be attributed to the diversity in decision-making in many cases due to a lack of knowledge and the buildup of experience that is needed to accurately diagnose endodontic cases [28, 29]. In contrast, our results documented a high degree of inter- and intra-observer agreement with the use of CBCT images; this may be explained by Patel et a [30] who demonstrated that examiner expertise had no effect on the interpretation of CBCT images. Voxel size is among the important factors that affect the diagnostic features of CBCT. In our study, there was no significant difference between the two voxel sizes (0.2 and 0.4mm³) in detecting the perforation, and this came in agreement with the previous studies [15, 16]. It is commonly believed that the clinical applications of high-resolution mode would depend on the clinical requirements and the acceptability from the point of view of radiation doses. The ALARA rule states that the ideal voxel size should produce the least amount of patient’s radiation dose and adequate image quality [31]. ALARA and radiation exposure must be taken into consideration while making CBCT [32, 33]. The high results gained by CBCT in detecting strip perforations in root-filled teeth in our study can be to some extent explained by the presence of filling material penetrating into the perforation [34]. It is probable that the lateral condensation of the gutta-percha in the current study caused the filling material, particularly the sealer, to penetrate the perforation defect, improving the detection of strip perforations radiographically. Additionally, our findings demonstrated that there was no significant difference in detecting strip perforation in teeth with and without obturation. Under the circumstances of this in vitro study, both PRs and CBCT scans were very accurate in identifying perforations in root-filled teeth; however, CBCT scans were much more accurate than PRs in identifying strip perforations. Therefore, the use of CBCT is advantageous and helpful in root perforation diagnostic procedures where it is not possible to detect the perforation based on PRs or clinical examination. The diagnostic effectiveness of CBCT scanning for perforation identification in clinical practice may be lower than that seen in this in vitro experiment. This in vitro model study has several limitations; dry human mandibles, soft tissue simulations, and artificially produced root perforations in this investigation did not produce images of the same quality as those in a clinical situation. The risk of measurement bias and overestimation may be increased by distinguishable characteristics of perforations that are produced artificially, a lack of patient movement, and inherent differences between artificial and natural settings [21]. Also, there are patient-related variables that can influence the clinical diagnosis of root perforations such as patient movement during scanning, beam-hardening artifacts produced by radiopaque materials, and high-density adjacent structures, such as metal posts, root canal filling materials, and restorations [35]. Future clinical trials using of PR and low dose CBCT in detection of root perforation should be performed.

Under the conditions of this in vitro study, regardless the observer specialty and when the images observed individually, the CBCT is superior to PR for detection of strip root perforation. Considering our results and the ALARA rule, a 0.4 mm³ voxel size has the same image quality and resolution as a 0.2 mm³ voxel size, so the lower radiation dose can therefore be used to minimize patient radiation exposure. Endodontist specialists showed a higher significance difference in the detection of root perforation using PR compared to general practitioners and radiologists, the accuracy for detection of the perforation when the two endodontist specialists observed the PR images was equal to CBCT accuracy.

Acknowledgement:

The authors would like to thank department of endodontics, faculty of dentistry, kafrelsheikh University, Egypt, and department of oral medicine, periodontology, diagnosis and oral radiology, kafrelsheikh university and Mansoura University, Egypt, for their participation in this research.

Conflict of interest: The authors declare no competing interests.

Author contribution:

Hisham Abada contributed to the study’s conception and design (lead author), Dana El gemaie wrote the discussion part of the article, Mohamed El Shreif collected the data, draw graphs and tables and Nour Hatata wrote the radiological part of the study. All authors contributed to data interpretation, drafting, and revision of the manuscript and gave final approval.

Ethics approval: Written ethical approval was obtained from the Institutional ethical board (no. KFSIRB200-1).

Funding: The authors received no specific funding for this article.

Pérez-Heredia M, Ferrer-Luque CM, Bravo M, et al (2017) cone-beam computed tomographic study of root anatomy and canal configuration of molars in a spanish population. J Endod 43:1511–1516. https://doi.org/10.1016/j.joen.2017.03.026
Mahmoudi E, Madani Z, Moudi E, et al (2019) diagnostic accuracy of high resolution cone-beam computed tomography and standard mode cone-beam computed tomography in internal root resorption. Iran Endod J 14:211–215. https://doi.org/10.22037/IEJ.V14I3.25005
Branco-de-Almeida LS, Velsko IM, de Oliveira ICV, et al (2023) impact of treatment on host responses in young individuals with periodontitis. J Dent Res. https://doi.org/10.1177/00220345221148161
Kim S, Kratchman S (2006) Modern endodontic surgery concepts and practice: a review. J Endod 32:601–23. https://doi.org/10.1016/j.joen.2005.12.010
Moshonov J, Michaeli E, Nahlieli O (2009) Endoscopic root canal treatment. undefined
Gordon MPJ, Chandler NP (2004) Electronic apex locators. Int Endod J 37:425–37. https://doi.org/10.1111/j.1365-2591.2004.00835.x
Fuss Z, Assooline LS, Kaufman AY (1996) Determination of location of root perforations by electronic apex locators. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 82:324–9. https://doi.org/10.1016/s1079-2104(96)80361-1
Shemesh H, van Soest G, Wu M-K, et al (2007) The ability of optical coherence tomography to characterize the root canal walls. J Endod 33:1369–73. https://doi.org/10.1016/j.joen.2007.06.022
Adel M, Tofangchiha M, Yeganeh LAB, et al (2016) diagnostic accuracy of cone-beam computed tomography and conventional periapical radiography in detecting strip root perforations. J Int Oral Heal 8:75–79
Safi Y, Ghaedsharaf S, Aziz A, et al (2017) effect of field of view on detection of external root resorption in cone-beam computed tomography. Iran Endod J 12:179–184. https://doi.org/10.22037/iej.2017.35
Venskutonis T, Juodzbalys G, Nackaerts O, Mickevicienė L (2013) Influence of voxel size on the diagnostic ability of cone-beam computed tomography to evaluate simulated root perforations. Oral Radiol 29:151–159. https://doi.org/10.1007/s11282-013-0125-5
Özer SY (2011) Detection of vertical root fractures by using cone beam computed tomography with variable voxel sizes in an in vitro model. J Endod 37:75–9. https://doi.org/10.1016/j.joen.2010.04.021
Safi Y, Hosseinpour S, Aziz A, et al (2016) Effect of amperage and field of view on detection of vertical root fracture in teeth with intracanal posts. Iran Endod J 11:202–7. https://doi.org/10.7508/iej.2016.03.011
Landis JR, Koch GG (1977) The measurement of observer agreement for categorical data. Biometrics 33:159. https://doi.org/10.2307/2529310
Koç C, Sönmez G, Yılmaz F, et al (2018) Comparison of the accuracy of periapical radiography with CBCT taken at 3 different voxel sizes in detecting simulated endodontic complications: An ex vivo study. Dentomaxillofacial Radiol 47:1–9. https://doi.org/10.1259/dmfr.20170399
Afkhami F, Ghoncheh Z, Khadiv F, et al (2021) How does voxel size of cone-beam computed tomography effect accurate detection of root strip perforations. Iran Endod J 16:43–48. https://doi.org/10.22037/iej.v16i1.25145
Skidmore AE, Bjorndal AM (1971) Root canal morphology of the human mandibular first molar. Oral Surgery, Oral Med Oral Pathol 32:778–784. https://doi.org/10.1016/0030-4220(71)90304-5
Berutti E, Fedon G (1992) Thickness of cementum/dentin in mesial roots of mandibular first molars. J Endod 18:545–548. https://doi.org/10.1016/S0099-2399(06)81211-2
Khojastepour L, Moazami F, Babaei M, Forghani M (2015) Assessment of root perforation within simulated internal resorption cavities using cone-beam computed tomography. J Endod 41:1520–1523. https://doi.org/10.1016/j.joen.2015.04.015
Bender IB, Seltzer S (2003) Roentgenographic and direct observation of experimental lesions in bone: I. J Endod 29:702–706. https://doi.org/10.1097/00004770-200311000-00005
Venskutonis T, Juodzbalys G, Nackaerts O, Mickevicienė L (2013) Influence of voxel size on the diagnostic ability of cone-beam computed tomography to evaluate simulated root perforations. Oral Radiol 29:151–159. https://doi.org/10.1007/s11282-013-0125-5
Patel S, Dawood A, Whaites E, Pitt Ford T (2009) New dimensions in endodontic imaging: part 1. Conventional and alternative radiographic systems. Int Endod J 42:447–462. https://doi.org/10.1111/J.1365-2591.2008.01530.X
Patel S (2009) New dimensions in endodontic imaging: Part 2. Cone beam computed tomography. Int Endod J 42:463–475. https://doi.org/10.1111/J.1365-2591.2008.01531.X
Durack C, Patel S, Davies J, et al (2011) Diagnostic accuracy of small volume cone beam computed tomography and intraoral periapical radiography for the detection of simulated external inflammatory root resorption. Int Endod J 44:136–147. https://doi.org/10.1111/j.1365-2591.2010.01819.x
Haiter-Neto F, Wenzel A, Gotfredsen E (2008) Diagnostic accuracy of cone beam computed tomography scans compared with intraoral image modalities for detection of caries lesions. Dentomaxillofacial Radiol 37:18–22. https://doi.org/10.1259/dmfr/87103878
Haghanifar S, Moudi E, Mesgarani A, et al (2014) A comparative study of cone-beam computed tomography and digital periapical radiography in detecting mandibular molars root perforations. Imaging Sci Dent 44:115–119. https://doi.org/10.5624/isd.2014.44.2.115
Shemesh H, Cristescu RC, Wesselink PR, Wu MK (2011) The use of cone-beam computed tomography and digital periapical radiographs to diagnose root perforations. J Endod 37:513–516. https://doi.org/10.1016/j.joen.2010.12.003
Burns LE, Visbal LD, Kohli MR, et al (2018) Long-term evaluation of treatment planning decisions for nonhealing endodontic cases by different groups of practitioners. J Endod 44:226–232. https://doi.org/10.1016/j.joen.2017.09.004
Kvist T (2001) Endodontic retreatment. Aspects of decision making and clinical outcome.
Patel S, Dawood A, Wilson R, et al (2009) The detection and management of root resorption lesions using intraoral radiography and cone beam computed tomography - an in vivo investigation. Int Endod J 42:831–838. https://doi.org/10.1111/J.1365-2591.2009.01592.X
Alawaji Y, MacDonald DS, Giannelis G, Ford NL (2018) Optimization of cone beam computed tomography image quality in implant dentistry. Clin Exp Dent Res 4:268–278. https://doi.org/10.1002/cre2.141
Loubele M, Bogaerts R, Van Dijck E, et al (2009) Comparison between effective radiation dose of CBCT and MSCT scanners for dentomaxillofacial applications. Eur J Radiol 71:461–468. https://doi.org/10.1016/j.ejrad.2008.06.002
Ludlow JB, Ivanovic M (2008) Comparative dosimetry of dental CBCT devices and 64-slice CT for oral and maxillofacial radiology. Oral Surgery, Oral Med Oral Pathol Oral Radiol Endodontology 106:106–114. https://doi.org/10.1016/j.tripleo.2008.03.018
Carvalho-Sousa B, Almeida-Gomes F, Borba Carvalho PR, et al (2010) Filling lateral canals: evaluation of different filling techniques. Eur J Dent 04:251–256. https://doi.org/10.1055/s-0039-1697836
Katsumata A, Hirukawa A, Noujeim M, et al (2006) Image artifact in dental cone-beam CT. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 101:652–657. https://doi.org/10.1016/J.TRIPLEO.2005.07.027

TABLE 1. Agreement and kappa values for intra-observer agreement in the detection of perforations between the same observers, for non-obturated and obturated teeth

Observers		Radiograph type	Non-obturated teeth		Obturated teeth
Observers		Radiograph type	Agreement (%)	Value	Agreement (%)	Value
Generarl Practitioners		PR	75	0.167	41.667	0.333
	1^st	CBCT-0.2	91.667	0.833	83.33	0.667
		CBCT-0.4	66.667	0.351	75	0.5
		PR	66.667	0.25	66.667	0.333
	2^nd	CBCT-0.2	83.333	0.657	100	1
		CBCT-0.4	100	1	100	1
Endodontists		PR	66.667	0.33	100	1
	1^st	CBCT-0.2	100	1	100	1
		CBCT-0.4	100	1	100	1
		PR	100	1	100	1
	2^nd	CBCT-0.2	100	1	100	1
		CBCT-0.4	91.667	0.833	100	1
Radiologists		PR	75	0.437	75	0.5
	1^st	CBCT-0.2	83.333	0.667	75	0.5
		CBCT-0.4	58.333	0.167	66.667	0.333
		PR	83.333	0.625	58.333	0.8
	2^nd	CBCT-0.2	91.667	0.833	75	0.5
		CBCT-0.4	100	1	83.333	0.667

TABLE 2. Kappa values for inter-observer agreement in the detection of perforations between the two observers (1^st and 2^nd observers), for non-obturated and obturated teeth.

Film type	Observers	Non-obturated teeth		Obturated teeth
Film type	Observers	Agreament (%)	Value	Agreament (%)	Value
PR	1^st VS 2^nd General practitioner	58.333	0.167	83.333	0.667
	1^st VS 2^nd Endodontist	66.667	0.333	100	1
	1^st VS 2^nd Radiologist	75	0.4	66.667	0.111
CBCT-0.2	1^st VS 2^nd General practitioner	100	1	100	1
	1^st VS 2^nd Endodontist	100	1	100	1
	1^st VS 2^nd Radiologist	100	1	91.667	0.833
CBCT-0.4	1^st VS 2^nd General practitioner	91.667	0.833	100	1
	1^st VS 2^nd Endodontist	100	1	100	1
	1^st VS 2^nd Radiologist	100	1	100	1

TABLE 3. AUC values, sensitivity, specificity, PPV, and NPV for all observers and image sets for non-obturated teeth

observer		film type	AUC values	SEa	Asymptotic sig.b	Asymptotic 95% confidence interval		Se (%)	Sp (%)	PPV (%)	NPV (%)
observer		film type	AUC values	SEa	Asymptotic sig.b	Lower bound	Upper bound	Se (%)	Sp (%)	PPV (%)	NPV (%)
Generarl Practitioners	1^st	PR	0.583	0.171	0.631	0.248	0.919	66.667	50	57.142	60
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	0.917	0.096	0.016	0.729	1	83.333	100	100	85.7142
	2^nd	PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Endodontists	1^st	PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
	2^nd	PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Radiologists	1^st	PR	0.75	0.15	0.15	0.456	1	100	50	66.666	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
	2^nd	PR	0.667	0.164	0.337	0.346	0.987	83.333	50	62.5	75
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Generarl Practitioners		PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Endodontists		PR	1	0	0.004	1	1	100	100	100	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Radiologists		PR	0.917	0.096	0.016	0.729	1	100	83.333	85.714	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100

TABLE 4. AUC values, sensitivity, specificity, PPV, an

observer		film type	AUC values	SEa	Asymptotic sig.b	Asymptotic 95% confidence interval		Se (%)	Sp (%)	PPV (%)	NPV (%)
observer		film type	AUC values	SEa	Asymptotic sig.b	Lower bound	Upper bound	Se (%)	Sp (%)	PPV (%)	NPV (%)
Generarl Practitioners	1^st	PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
	2^nd	PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Endodontists	1^st	PR	1	0	0.004	1	1	100	100	100	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
	2^nd	PR	1	0	0.004	1	1	100	100	100	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Radiologists	1^st	PR	0.75	0.15	0.15	0.456	1	100	50	66.666	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
	2^nd	PR	0.75	0.15	0.15	0.456	1	100	50	66.666	100
		CBCT-0.2	0.917	0.096	0.016	0.729	1	100	83.333	85.714	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Generarl Practitioners		PR	0.833	0.129	0.055	0.58	1	83.333	83.333	83.333	83.333
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Endodontists		PR	1	0	0.004	1	1	100	100	100	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100
Two Radiologists		PR	0.917	0.096	0.016	0.729	1	100	83.333	85.714	100
		CBCT-0.2	1	0	0.004	1	1	100	100	100	100
		CBCT-0.4	1	0	0.004	1	1	100	100	100	100

No competing interests reported.

Accuracy of Cone-Beam Computed Tomography in Detection of Root Perforations Using Different Voxel Sizes in Comparison to Digital Periapical Radiograph, An in vitro study

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Results

Discussion

Conclusion

Declarations

References

Tables

Additional Declarations

Status:

Version 1