- Email: [email protected]
Diagnostic Accuracy of Periapical Radiography and Cone-beam Computed Tomography in Identifying Root Canal Configuration of Human Premolars
Basic Research—Technology
Diagnostic Accuracy of Periapical Radiography and Cone-beam Computed Tomography in Identifying Root Canal Configuration of Human Premolars Thiago Oliveira Sousa, DDS, MSc, PhD,* Francisco Haiter-Neto, DDS, MSc, PhD,* Eduarda Helena Leandro Nascimento, DDS, MSc,* Leonardo Vieira Peroni, DDS, MSc,* Deborah Queiroz Freitas, DDS, MSc, PhD,* and Bassam Hassan, BDS, MSc, PhD† Abstract Introduction: The aim of this study was to assess the diagnostic accuracy of periapical radiography (PR) and cone-beam computed tomographic (CBCT) imaging in the detection of the root canal configuration (RCC) of human premolars. Methods: PR and CBCT imaging of 114 extracted human premolars were evaluated by 2 oral radiologists. RCC was recorded according to Vertucci’s classification. Micro–computed tomographic imaging served as the gold standard to determine RCC. Accuracy, sensitivity, specificity, and predictive values were calculated. The Friedman test compared both PR and CBCT imaging with the gold standard. Results: CBCT imaging showed higher values for all diagnostic tests compared with PR. Accuracy was 0.55 and 0.89 for PR and CBCT imaging, respectively. There was no difference between CBCT imaging and the gold standard, whereas PR differed from both CBCT and micro– computed tomographic imaging (P < .0001). CBCT imaging was more accurate than PR for evaluating different types of RCC individually. Canal configuration types III, VII, and ‘‘other’’ were poorly identified on CBCT imaging with a detection accuracy of 50%, 0%, and 43%, respectively. With PR, all canal configurations except type I were poorly visible. Conclusions: PR presented low performance in the detection of RCC in premolars, whereas CBCT imaging showed no difference compared with the gold standard. Canals with complex configurations were less identifiable using both imaging methods, especially PR. (J Endod 2017;-:1–4)
Key Words Anatomy, cone-beam computed tomography, dental pulp cavity, endodontics, radiography
T
he anatomy of root caSignificance nals has long been Adequate diagnosis and treatment of root canals investigated in the literature are challenging because of their anatomic since pioneer in vitro complexity. Periapical radiography and CBCT imstudies showed how comaging have an important role in daily practice. The plex the internal tooth study aim was to evaluate their effectiveness in morphology can be (1). identifying root canal configurations in extracted Indeed, knowledge of root premolars. canal configuration (RCC) is crucial to achieve treatment success because it assists in drawing up an appropriate plan and, consequently, avoids possible technical failures at all stages of treatment (2–4). Periapical radiography (PR) is the most widely adopted method to evaluate root canal anatomy in clinical practice (5). This simple technique provides complementary information at a relatively low cost and radiation dose. However, despite its widespread use, PR fails to depict the complex anatomic configuration of teeth because of image overlapping inherent to conventional 2-dimensional radiography. Cone-beam computed tomographic (CBCT) imaging provides high-quality, accurate, 3-dimensional (3D) representations of hard tissues, resulting in a more accurate diagnosis of many conditions (6). Although it is not indicated for the initial evaluation of dental morphology, CBCT imaging may be considered when it has been decided that radiographic images are yielding limited information and that further details are required for diagnosis and treatment planning (7). Many studies have used both PR and CBCT imaging to identify RCC in different populations with no gold standard or have simply used CBCT imaging as the reference method (8–11). However, the assumption that CBCT scanning is accurate enough to diagnose RCC without comparing it with a gold standard can underestimate the complexity of RCC. Therefore, the precision and biases of RCC frequency-based studies cannot be estimated because of the lack of knowledge of the accuracy of CBCT imaging in detecting RCC. Micro–computed tomographic (mCT) imaging is often used as the gold standard because of its high spatial resolution, which allows endodontic features to be evaluated in detail without the destruction of samples (12). To the best of our knowledge, the efficacy of CBCT imaging and PR (which is the most common imaging modality) in
From the *Department of Oral Diagnosis, Piracicaba Dental School, University of Campinas, Campinas, S~ao Paulo, Brazil; and †Department of Oral Implantology and Prosthetic Dentistry, Academic Center for Dentistry Amsterdam, Universiteit van Amsterdam and Vrije Universiteit, The Netherlands. Address requests for reprints to Dr Thiago Oliveira Sousa, Programa de Pos-Graduac¸~ao em Radiologia Odontologica, Avenida Limeira, n 901, Bairro Arei~ao, CP 52, Piracicaba, SP, Brazil, CEP: 13414-903. E-mail address: [email protected] 0099-2399/$ - see front matter Copyright ª 2017 American Association of Endodontists. http://dx.doi.org/10.1016/j.joen.2017.02.021
JOE — Volume -, Number -, - 2017
Identifying RCC of Human Premolars
1
Basic Research—Technology identifying RCC of human premolars supported by a gold standard method has not been investigated yet. This study aimed to assess the diagnostic accuracy of PR and CBCT images in detecting RCC of human premolars using mCT imaging as the gold standard.
imaging modality, images were re-evaluated by consensus of the 2 evaluators. For each imaging modality, a second evaluation was performed under the same conditions with 20% of the sample to assess the method’s reproducibility.
Materials and Methods
Statistical Analysis Data were analyzed using SPSS software for Windows (Version 22; SPSS Corp, Chicago, IL). Values of accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were obtained. A comparison among PR, CBCT, and mCT findings was performed using the Friedman test with a significance level (a) of 5%. The kappa test was performed to assess intraobserver reliability.
Data Collection After local ethics committee approval (protocol number 154/ 2015), 114 single- and 2-rooted extracted human premolars were collected, cleaned, and disinfected. Maxillary (11 first and 45 second) and mandibular (34 first and 24 second) premolars were included. Exclusion criteria consisted of teeth with root canal treatment, calcification, incomplete root formation, resorptive lesions, or fractures. To simulate clinical conditions, the teeth were placed in a dry mandible in the premolar region covered with wax. Periapical radiographs were obtained with the paralleling technique using a dental X-ray device (Focus; Instrumentarium, Tuusula, Finland) and the VistaScan intraoral digital system (D€urr Dental, Beitigheim-Bissinger, Germany) operating at 7 mA, 70 kVp, and 0.06 seconds. CBCT scans were performed using the 3D Accuitomo device (J Morita Manufacturing, Kyoto, Japan) with a high-resolution protocol (field of view: 4 4, a voxel size of 0.08 mm, 90 kVp, and 5 mA). Before CBCT scanning, the mounted dry mandible was submerged in water in a 12 7.5 7 cm plastic container to simulate the natural effect of radiation attenuation and scattering produced by soft tissues (13). The teeth were then scanned on a Skyscan 1174 mCT unit (Bruker, Kontich, Belgium) using the following settings: 50 kV, 800 mA, a 15.91-mm voxel size, a 1.0-mm aluminum filter, a rotation step 0.4 , and 4 frames. Image Assessment The data set (PR and CBCT images) was randomized and evaluated independently by 2 oral radiologists with more than 5 years of experience in diagnostic imaging. First, RCC was assessed in periapical radiographs using ImageJ software (National Institutes of Health, Bethesda, MD), and each main canal was classified according to Vertucci (14) (Fig. 1). Root canals with a configuration other than the ones present in Vertucci’s classification were assigned as ‘‘other.’’ CBCT images were randomized and evaluated using CS 3D Imaging Version 3.5.7 software (Carestream Health Inc, Rochester, NY) dynamically in the multiplanar reconstruction mode. Zoom, brightness, and contrast tools were available to be used for both periapical and CBCT evaluations. After PR and CBCT evaluations, mCT images were analyzed by 2 other independent evaluators using CTAn software (v.1.14.4.1; Bruker, Kontich, Belgium) to obtain the gold standard. In case of disagreement in any
Results Vertucci’s types of RCCs identified after mCT analysis were type I (49.18%), type II (2.46%), type III (3.28%), type IV (7.38%), type V (12.29%), type VI (4.1%), and type VII (2.46%), and 18.85% of the canals were categorized as ‘‘other.’’ There was no type VIII configuration in the sample. CBCT imaging showed consistent diagnostic results and higher accuracy than PR (Table 1). PR showed low diagnostic values, except specificity. The Friedman test found no difference between CBCT imaging and the gold standard, whereas PR significantly differed from both CBCT and mCT imaging (P < .0001) (Fig. 2). Table 2 details the comparison of PR and CBCT imaging with the gold standard, considering different types of RCC individually. Overall, CBCT identification was more accurate than PR. Types III, VII, and ‘‘other’’ had the poorest detection scores using CBCT imaging (50%, 0%, and 43%, respectively). In PR evaluation, except for type I, all other higher-complexity types presented critical identification; only 2 of 66 canals were correctly identified. In 77% of these mistaken answers, type I was the most common reported answer. Kappa values for intraobserver reproducibility for PR and CBCT imaging were considered almost perfect agreement and substantial agreement (0.83 and 0.78, respectively) according to the classification of Landis and Kock (15).
Discussion This study is the first in which PR and CBCT imaging were tested and compared regarding the correct identification of RCC in premolars using mCT imaging as the gold standard. CBCT imaging was more accurate than PR in the assessment of 122 root canals and provided better results in all diagnostic tests. When confronting the data from the 3 imaging modalities, CBCT imaging did not differ from mCT imaging, whereas PR significantly differed from both.
Figure 1. Vertucci’s classification of root canal configuration: type I, a single canal; type II, 2 canals that converge near the apex; type III, a canal that is divided in 2 but converges again near the apex; type IV, 2 independent canals; type V, a canal that is divided in 2 before the apex; type VI, 2 canals that converge within the root and are divided again into 2 canals; type VII, a canal that is divided and then converges within the root and is again divided into 2 before the apex; and type VIII, 3 independent canals.
2
Sousa et al.
JOE — Volume -, Number -, - 2017
Basic Research—Technology TABLE 1. Accuracy, Sensitivity, Specificity, Positive Predictive Value (PPV), and Negative Predictive Value (NPV) for the Identification of Root Canal Configuration Using Periapical Radiography (PR) and Cone-beam Computed Tomographic (CBCT) Imaging Imaging method PR CBCT
Accuracy
Sensitivity
Specificity
PPV
NPV
0.55 0.89
0.18 0.79
0.93 0.98
0.73 0.98
0.52 0.82
To validate CBCT imaging as a tool to explore root canal anatomy, a previous study (16) used 9 extracted teeth and histologic sections, but it was not conclusive, suggesting further investigation on the diagnostic capacity of CBCT imaging is needed. Despite the fact that the present study used a high-resolution protocol with a voxel size of 0.08 mm, CBCT imaging failed to show some RCCs, such as types III, VII, and ‘‘other.’’ However, considering the overall results, CBCT imaging showed a high degree of accuracy (89%) in identifying the complexity of RCC. It is worth emphasizing that although the CBCT radiation dose is fairly low compared with multidetector computed tomographic imaging, it is still considerably higher than the PR dose. This implies that CBCT indication should never be done uncritically but only in cases in which possible risks to radiation exposure are strongly justified by the diagnostic improvement (17). Most of the type I canals were correctly identified using PR, whereas all the other types were misdiagnosed as type I in almost all cases. This is explained by the 2-dimensional condition of PR images
and consequently the superposition of the canal features, which may have affected the results for complex RCCs. When investigating root canal systems with PR images, Matherne et al (18) reported that the observer failed to identify at least 1 root canal in approximately 40% of teeth compared with CBCT imaging. Their failure index for PR is slightly lower than in the current study; nevertheless, their gold standard was CBCT imaging analyzed by only 1 observer. In a previous study (12) using PR, CBCT, and mCT imaging to assess anatomic variations of root canals (including RCC) in 40 mandibular incisors, high accuracy was found for both techniques. Despite having comparable results for Vertucci type I canals, the study was limited to only types I and III, which does not allow an unbiased correlation with our study when considering all groups analyzed. Evaluating the RCC of first mandibular premolars using CBCT images, Zhang et al (19) correctly identified 98% of type I canals and 100% of type III and V canals but could not identify type VII. In the present study, results for type I and type VII canals presented an equal ratio of correct identification (98% and 0%, respectively), whereas for types III and V the results were slightly lower. This can be explained by the low number of cases of these specific types of RCC and by the fact that they used a voxel size of 125 mm and only 1 observer, whereas in the present study a 8-mm voxel size and 2 observers were used. It is known that voxel size (which influences the contrast and spatial resolution) plays an important role in the ability of CBCT imaging to display fine details (20–24). Therefore, the use of a high-resolution protocol with a voxel size as small as possible should be preferred in these cases (22, 24). Although when using PR complex RCC was misdiagnosed as type I, distinct mistakes were also made when using CBCT imaging. Type III
Figure 2. mCT 3-dimensional reconstruction (1), PR (2), and CBCT coronal reconstruction (3) with respective axial reconstructions in different levels (a–d) of the same root canal: an example of type I RCC. Images 4, 5, and 6a–d illustrate an example of a type VI canal (with a higher complexity of RCC than type I).
JOE — Volume -, Number -, - 2017
Identifying RCC of Human Premolars
3
Basic Research—Technology TABLE 2. Identification of Root Canal Configuration Types (N = 122) in Periapical Radiography (PR) and Cone-beam Computed Tomographic (CBCT) Imaging Compared with the Gold Standard (Micro–computed Tomographic [mCT] Imaging) Imaging method PR Type I Type II Type III Type IV Type V Type VI Type VII Type VIII Other Total CBCT Type I Type II Type III Type IV Type V Type VI Type VII Type VIII Other Total
mCT imaging I
II III IV
V
VI VII VIII Other Total
56 0 1 0 2 0 0 0 1 60
3 0 0 0 0 0 0 0 0 3
4 0 0 0 0 0 0 0 0 4
5 0 2 1 1 0 0 0 0 9
14 0 0 0 1 0 0 0 0 15
3 0 2 0 0 0 0 0 0 5
3 0 0 0 0 0 0 0 0 3
0 0 0 0 0 0 0 0 0 0
19 0 1 0 3 0 0 0 0 23
107 0 6 1 7 0 0 0 1 122
59 0 0 0 0 1 0 0 0 60
0 3 0 0 0 0 0 0 0 3
0 2 2 0 0 0 0 0 0 4
2 0 0 7 0 0 0 0 0 9
5 0 0 0 9 0 0 0 1 15
0 0 0 0 0 4 1 0 0 5
0 0 1 0 0 1 0 0 1 3
0 0 0 0 0 0 0 0 0 0
6 1 0 1 3 1 1 0 10 23
72 6 3 8 12 7 2 0 12 122
Bold numbers in the diagonal represent matching answers between mCT imaging and each of the analyzed methods. Numbers outside the diagonal represent wrong answers given by the evaluators.
was mistaken for type V in 2 of 4 cases, whereas 5 of 15 type V canals were mistaken for type I. Clinically, those misidentifications can result in a missed canal during instrumentation/obturation and can lead to unsatisfactory root canal treatment. In conclusion, CBCT imaging showed high accuracy in RCC identification, whereas PR presented poor results deviating significantly from the gold standard. These findings indicate that CBCT imaging contributed effectively to RCC identification. Contrarily, PR was inefficient to identify complex canal configurations. Although CBCT imaging appears to be the most effective imaging examination applicable clinically for identifying RCC, clinicians should always respect the indication criteria and be aware of the limitations of CBCT imaging in displaying the configurations in some cases.
Acknowledgments The authors thank CAPES Foundation (Ministry of Education of Brazil, Brasılia/DF–Brazil) for granting a scholarship to the first author during this research. The authors deny any conflicts of interest related to this study.
References
2. Siqueira JF Jr. Endodontic infections: concepts, paradigms, and perspectives. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;94:281–93. 3. Vertucci FJ. Root canal morphology and its relationship to endodontic procedures. Endod Topics 2005;10:3–29. 4. Yilmaz Z, Tuncel B, Serper A, et al. C-shaped root canal in a maxillary first molar: a case report. Int Endod J 2006;39:162–6. 5. Fan B, Gao Y, Fan W, et al. Identification of a C-shaped canal system in mandibular second molars-part II: the effect of bone image superimposition and intraradicular contrast medium on radiograph interpretation. J Endod 2008;34:160–5. 6. Scarfe WC, Levin MD, Gane D, et al. Use of cone beam computed tomography in endodontics. Int J Dent 2009;2009:634567. 7. Patel S, Durack C, Abella F, et al. Cone beam computed tomography in endodontics a review. Int Endod J 2015;48:3–15. 8. de Pablo OV, Estevez R, Peix Sanchez M, et al. Root anatomy and canal configuration of the permanent mandibular first molar: a systematic review. J Endod 2010;36: 1919–31. 9. Tian YY, Guo B, Zhang R, et al. Root and canal morphology of maxillary first premolars in a Chinese subpopulation evaluated using cone-beam computed tomography. Int Endod J 2012;45:996–1003. 10. Kottoor J, Albuquerque D, Velmurugan N, et al. Root anatomy and root canal configuration of human permanent mandibular premolars: a systematic review. Anat Res Int 2013;2013:254250. 11. Silva EJNL, Nejaim Y, Silva AV, et al. Evaluation of root canal configuration of mandibular molars in a Brazilian population by using cone-beam computed tomography: an in vivo study. J Endod 2013;39:849–52. 12. Paes Da Silva Ramos Fernandes LM, Rice D, Ordinola-Zapata R, et al. Detection of various anatomic patterns of root canals in mandibular incisors using digital periapical radiography, 3 cone-beam computed tomographic scanners, and microcomputed tomographic imaging. J Endod 2014;40:42–5. 13. Neves FS, Freitas DQ, Campos PS, et al. Evaluation of cone-beam computed tomography in the diagnosis of vertical root fractures: the influence of imaging modes and root canal materials. J Endod 2014;40:1530–6. 14. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral Med Oral Pathol 1984;58:589–99. 15. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Published by: International Biometric Society Stable 1977;33:159–74. 16. Michetti J, Maret D, Mallet JP, et al. Validation of cone beam computed tomography as a tool to explore root canal anatomy. J Endod 2010;36:1187–90. 17. Jaju PP, Jaju SP. Cone-beam computed tomography: time to move from ALARA to ALADA. Imaging Sci Dent 2015;45:263–5. 18. Matherne RP, Angelopoulos C, Kulild JC, et al. Use of cone-beam computed tomography to identify root canal systems in vitro. J Endod 2008;34:87–9. 19. Zhang D, Chen J, Lan G, et al. The root canal morphology in mandibular first premolars: a comparative evaluation of cone-beam computed tomography and microcomputed tomography. Clin Oral Investig 2016 May 16; http://dx.doi.org/10.1007/ s00784-016-1852-x [Epub ahead of print]. 20. Bauman R, Scarfe W, Clark SJ, et al. Ex vivo detection of mesiobuccal canals in maxillary molars using CBCT at four different isotropic voxel dimensions. Int Endod J 2011;44:752–8. 21. Hassan B, Metska ME, Ozok AR, et al. Comparison of five cone beam computed tomography systems for the detection of vertical root fractures. J Endod 2010;36: 126–9. 22. Kamburoglu K, Murat S, Kolsuz E, et al. Comparative assessment of subjective image quality of cross-sectional cone-beam computed tomography scans. J Oral Sci 2011; 53:501–8. 23. Liedke GS, Spin-Neto R, Vizzotto MB, et al. Diagnostic accuracy of cone-beam computed tomography sections with various thicknesses for detecting misfit between the tooth and restoration in metal-restored teeth. Oral Surg Oral Med Oral Pathol Oral Radiol 2015;120:131–7. 24. Th€onissen P, Ermer MA, Schmelzeisen R, et al. Sensitivity and specificity of cone beam computed tomography in thin bony structures in maxillofacial surgery – a clinical trial. J Craniomaxillofac Surg 2015;43:1284–8.
1. Okumura T. Anatomy of the root canals. J Am Dent Assoc 1927;14:632–9.
4
Sousa et al.
JOE — Volume -, Number -, - 2017
Diagnostic Accuracy of Periapical Radiography and Cone-beam Computed Tomography in Identifying Root Canal Configuration of Human Premolars Thiago Oliveira Sousa, DDS, MSc, PhD,* Francisco Haiter-Neto, DDS, MSc, PhD,* Eduarda Helena Leandro Nascimento, DDS, MSc,* Leonardo Vieira Peroni, DDS, MSc,* Deborah Queiroz Freitas, DDS, MSc, PhD,* and Bassam Hassan, BDS, MSc, PhD† Abstract Introduction: The aim of this study was to assess the diagnostic accuracy of periapical radiography (PR) and cone-beam computed tomographic (CBCT) imaging in the detection of the root canal configuration (RCC) of human premolars. Methods: PR and CBCT imaging of 114 extracted human premolars were evaluated by 2 oral radiologists. RCC was recorded according to Vertucci’s classification. Micro–computed tomographic imaging served as the gold standard to determine RCC. Accuracy, sensitivity, specificity, and predictive values were calculated. The Friedman test compared both PR and CBCT imaging with the gold standard. Results: CBCT imaging showed higher values for all diagnostic tests compared with PR. Accuracy was 0.55 and 0.89 for PR and CBCT imaging, respectively. There was no difference between CBCT imaging and the gold standard, whereas PR differed from both CBCT and micro– computed tomographic imaging (P < .0001). CBCT imaging was more accurate than PR for evaluating different types of RCC individually. Canal configuration types III, VII, and ‘‘other’’ were poorly identified on CBCT imaging with a detection accuracy of 50%, 0%, and 43%, respectively. With PR, all canal configurations except type I were poorly visible. Conclusions: PR presented low performance in the detection of RCC in premolars, whereas CBCT imaging showed no difference compared with the gold standard. Canals with complex configurations were less identifiable using both imaging methods, especially PR. (J Endod 2017;-:1–4)
Key Words Anatomy, cone-beam computed tomography, dental pulp cavity, endodontics, radiography
T
he anatomy of root caSignificance nals has long been Adequate diagnosis and treatment of root canals investigated in the literature are challenging because of their anatomic since pioneer in vitro complexity. Periapical radiography and CBCT imstudies showed how comaging have an important role in daily practice. The plex the internal tooth study aim was to evaluate their effectiveness in morphology can be (1). identifying root canal configurations in extracted Indeed, knowledge of root premolars. canal configuration (RCC) is crucial to achieve treatment success because it assists in drawing up an appropriate plan and, consequently, avoids possible technical failures at all stages of treatment (2–4). Periapical radiography (PR) is the most widely adopted method to evaluate root canal anatomy in clinical practice (5). This simple technique provides complementary information at a relatively low cost and radiation dose. However, despite its widespread use, PR fails to depict the complex anatomic configuration of teeth because of image overlapping inherent to conventional 2-dimensional radiography. Cone-beam computed tomographic (CBCT) imaging provides high-quality, accurate, 3-dimensional (3D) representations of hard tissues, resulting in a more accurate diagnosis of many conditions (6). Although it is not indicated for the initial evaluation of dental morphology, CBCT imaging may be considered when it has been decided that radiographic images are yielding limited information and that further details are required for diagnosis and treatment planning (7). Many studies have used both PR and CBCT imaging to identify RCC in different populations with no gold standard or have simply used CBCT imaging as the reference method (8–11). However, the assumption that CBCT scanning is accurate enough to diagnose RCC without comparing it with a gold standard can underestimate the complexity of RCC. Therefore, the precision and biases of RCC frequency-based studies cannot be estimated because of the lack of knowledge of the accuracy of CBCT imaging in detecting RCC. Micro–computed tomographic (mCT) imaging is often used as the gold standard because of its high spatial resolution, which allows endodontic features to be evaluated in detail without the destruction of samples (12). To the best of our knowledge, the efficacy of CBCT imaging and PR (which is the most common imaging modality) in
From the *Department of Oral Diagnosis, Piracicaba Dental School, University of Campinas, Campinas, S~ao Paulo, Brazil; and †Department of Oral Implantology and Prosthetic Dentistry, Academic Center for Dentistry Amsterdam, Universiteit van Amsterdam and Vrije Universiteit, The Netherlands. Address requests for reprints to Dr Thiago Oliveira Sousa, Programa de Pos-Graduac¸~ao em Radiologia Odontologica, Avenida Limeira, n 901, Bairro Arei~ao, CP 52, Piracicaba, SP, Brazil, CEP: 13414-903. E-mail address: [email protected] 0099-2399/$ - see front matter Copyright ª 2017 American Association of Endodontists. http://dx.doi.org/10.1016/j.joen.2017.02.021
JOE — Volume -, Number -, - 2017
Identifying RCC of Human Premolars
1
Basic Research—Technology identifying RCC of human premolars supported by a gold standard method has not been investigated yet. This study aimed to assess the diagnostic accuracy of PR and CBCT images in detecting RCC of human premolars using mCT imaging as the gold standard.
imaging modality, images were re-evaluated by consensus of the 2 evaluators. For each imaging modality, a second evaluation was performed under the same conditions with 20% of the sample to assess the method’s reproducibility.
Materials and Methods
Statistical Analysis Data were analyzed using SPSS software for Windows (Version 22; SPSS Corp, Chicago, IL). Values of accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were obtained. A comparison among PR, CBCT, and mCT findings was performed using the Friedman test with a significance level (a) of 5%. The kappa test was performed to assess intraobserver reliability.
Data Collection After local ethics committee approval (protocol number 154/ 2015), 114 single- and 2-rooted extracted human premolars were collected, cleaned, and disinfected. Maxillary (11 first and 45 second) and mandibular (34 first and 24 second) premolars were included. Exclusion criteria consisted of teeth with root canal treatment, calcification, incomplete root formation, resorptive lesions, or fractures. To simulate clinical conditions, the teeth were placed in a dry mandible in the premolar region covered with wax. Periapical radiographs were obtained with the paralleling technique using a dental X-ray device (Focus; Instrumentarium, Tuusula, Finland) and the VistaScan intraoral digital system (D€urr Dental, Beitigheim-Bissinger, Germany) operating at 7 mA, 70 kVp, and 0.06 seconds. CBCT scans were performed using the 3D Accuitomo device (J Morita Manufacturing, Kyoto, Japan) with a high-resolution protocol (field of view: 4 4, a voxel size of 0.08 mm, 90 kVp, and 5 mA). Before CBCT scanning, the mounted dry mandible was submerged in water in a 12 7.5 7 cm plastic container to simulate the natural effect of radiation attenuation and scattering produced by soft tissues (13). The teeth were then scanned on a Skyscan 1174 mCT unit (Bruker, Kontich, Belgium) using the following settings: 50 kV, 800 mA, a 15.91-mm voxel size, a 1.0-mm aluminum filter, a rotation step 0.4 , and 4 frames. Image Assessment The data set (PR and CBCT images) was randomized and evaluated independently by 2 oral radiologists with more than 5 years of experience in diagnostic imaging. First, RCC was assessed in periapical radiographs using ImageJ software (National Institutes of Health, Bethesda, MD), and each main canal was classified according to Vertucci (14) (Fig. 1). Root canals with a configuration other than the ones present in Vertucci’s classification were assigned as ‘‘other.’’ CBCT images were randomized and evaluated using CS 3D Imaging Version 3.5.7 software (Carestream Health Inc, Rochester, NY) dynamically in the multiplanar reconstruction mode. Zoom, brightness, and contrast tools were available to be used for both periapical and CBCT evaluations. After PR and CBCT evaluations, mCT images were analyzed by 2 other independent evaluators using CTAn software (v.1.14.4.1; Bruker, Kontich, Belgium) to obtain the gold standard. In case of disagreement in any
Results Vertucci’s types of RCCs identified after mCT analysis were type I (49.18%), type II (2.46%), type III (3.28%), type IV (7.38%), type V (12.29%), type VI (4.1%), and type VII (2.46%), and 18.85% of the canals were categorized as ‘‘other.’’ There was no type VIII configuration in the sample. CBCT imaging showed consistent diagnostic results and higher accuracy than PR (Table 1). PR showed low diagnostic values, except specificity. The Friedman test found no difference between CBCT imaging and the gold standard, whereas PR significantly differed from both CBCT and mCT imaging (P < .0001) (Fig. 2). Table 2 details the comparison of PR and CBCT imaging with the gold standard, considering different types of RCC individually. Overall, CBCT identification was more accurate than PR. Types III, VII, and ‘‘other’’ had the poorest detection scores using CBCT imaging (50%, 0%, and 43%, respectively). In PR evaluation, except for type I, all other higher-complexity types presented critical identification; only 2 of 66 canals were correctly identified. In 77% of these mistaken answers, type I was the most common reported answer. Kappa values for intraobserver reproducibility for PR and CBCT imaging were considered almost perfect agreement and substantial agreement (0.83 and 0.78, respectively) according to the classification of Landis and Kock (15).
Discussion This study is the first in which PR and CBCT imaging were tested and compared regarding the correct identification of RCC in premolars using mCT imaging as the gold standard. CBCT imaging was more accurate than PR in the assessment of 122 root canals and provided better results in all diagnostic tests. When confronting the data from the 3 imaging modalities, CBCT imaging did not differ from mCT imaging, whereas PR significantly differed from both.
Figure 1. Vertucci’s classification of root canal configuration: type I, a single canal; type II, 2 canals that converge near the apex; type III, a canal that is divided in 2 but converges again near the apex; type IV, 2 independent canals; type V, a canal that is divided in 2 before the apex; type VI, 2 canals that converge within the root and are divided again into 2 canals; type VII, a canal that is divided and then converges within the root and is again divided into 2 before the apex; and type VIII, 3 independent canals.
2
Sousa et al.
JOE — Volume -, Number -, - 2017
Basic Research—Technology TABLE 1. Accuracy, Sensitivity, Specificity, Positive Predictive Value (PPV), and Negative Predictive Value (NPV) for the Identification of Root Canal Configuration Using Periapical Radiography (PR) and Cone-beam Computed Tomographic (CBCT) Imaging Imaging method PR CBCT
Accuracy
Sensitivity
Specificity
PPV
NPV
0.55 0.89
0.18 0.79
0.93 0.98
0.73 0.98
0.52 0.82
To validate CBCT imaging as a tool to explore root canal anatomy, a previous study (16) used 9 extracted teeth and histologic sections, but it was not conclusive, suggesting further investigation on the diagnostic capacity of CBCT imaging is needed. Despite the fact that the present study used a high-resolution protocol with a voxel size of 0.08 mm, CBCT imaging failed to show some RCCs, such as types III, VII, and ‘‘other.’’ However, considering the overall results, CBCT imaging showed a high degree of accuracy (89%) in identifying the complexity of RCC. It is worth emphasizing that although the CBCT radiation dose is fairly low compared with multidetector computed tomographic imaging, it is still considerably higher than the PR dose. This implies that CBCT indication should never be done uncritically but only in cases in which possible risks to radiation exposure are strongly justified by the diagnostic improvement (17). Most of the type I canals were correctly identified using PR, whereas all the other types were misdiagnosed as type I in almost all cases. This is explained by the 2-dimensional condition of PR images
and consequently the superposition of the canal features, which may have affected the results for complex RCCs. When investigating root canal systems with PR images, Matherne et al (18) reported that the observer failed to identify at least 1 root canal in approximately 40% of teeth compared with CBCT imaging. Their failure index for PR is slightly lower than in the current study; nevertheless, their gold standard was CBCT imaging analyzed by only 1 observer. In a previous study (12) using PR, CBCT, and mCT imaging to assess anatomic variations of root canals (including RCC) in 40 mandibular incisors, high accuracy was found for both techniques. Despite having comparable results for Vertucci type I canals, the study was limited to only types I and III, which does not allow an unbiased correlation with our study when considering all groups analyzed. Evaluating the RCC of first mandibular premolars using CBCT images, Zhang et al (19) correctly identified 98% of type I canals and 100% of type III and V canals but could not identify type VII. In the present study, results for type I and type VII canals presented an equal ratio of correct identification (98% and 0%, respectively), whereas for types III and V the results were slightly lower. This can be explained by the low number of cases of these specific types of RCC and by the fact that they used a voxel size of 125 mm and only 1 observer, whereas in the present study a 8-mm voxel size and 2 observers were used. It is known that voxel size (which influences the contrast and spatial resolution) plays an important role in the ability of CBCT imaging to display fine details (20–24). Therefore, the use of a high-resolution protocol with a voxel size as small as possible should be preferred in these cases (22, 24). Although when using PR complex RCC was misdiagnosed as type I, distinct mistakes were also made when using CBCT imaging. Type III
Figure 2. mCT 3-dimensional reconstruction (1), PR (2), and CBCT coronal reconstruction (3) with respective axial reconstructions in different levels (a–d) of the same root canal: an example of type I RCC. Images 4, 5, and 6a–d illustrate an example of a type VI canal (with a higher complexity of RCC than type I).
JOE — Volume -, Number -, - 2017
Identifying RCC of Human Premolars
3
Basic Research—Technology TABLE 2. Identification of Root Canal Configuration Types (N = 122) in Periapical Radiography (PR) and Cone-beam Computed Tomographic (CBCT) Imaging Compared with the Gold Standard (Micro–computed Tomographic [mCT] Imaging) Imaging method PR Type I Type II Type III Type IV Type V Type VI Type VII Type VIII Other Total CBCT Type I Type II Type III Type IV Type V Type VI Type VII Type VIII Other Total
mCT imaging I
II III IV
V
VI VII VIII Other Total
56 0 1 0 2 0 0 0 1 60
3 0 0 0 0 0 0 0 0 3
4 0 0 0 0 0 0 0 0 4
5 0 2 1 1 0 0 0 0 9
14 0 0 0 1 0 0 0 0 15
3 0 2 0 0 0 0 0 0 5
3 0 0 0 0 0 0 0 0 3
0 0 0 0 0 0 0 0 0 0
19 0 1 0 3 0 0 0 0 23
107 0 6 1 7 0 0 0 1 122
59 0 0 0 0 1 0 0 0 60
0 3 0 0 0 0 0 0 0 3
0 2 2 0 0 0 0 0 0 4
2 0 0 7 0 0 0 0 0 9
5 0 0 0 9 0 0 0 1 15
0 0 0 0 0 4 1 0 0 5
0 0 1 0 0 1 0 0 1 3
0 0 0 0 0 0 0 0 0 0
6 1 0 1 3 1 1 0 10 23
72 6 3 8 12 7 2 0 12 122
Bold numbers in the diagonal represent matching answers between mCT imaging and each of the analyzed methods. Numbers outside the diagonal represent wrong answers given by the evaluators.
was mistaken for type V in 2 of 4 cases, whereas 5 of 15 type V canals were mistaken for type I. Clinically, those misidentifications can result in a missed canal during instrumentation/obturation and can lead to unsatisfactory root canal treatment. In conclusion, CBCT imaging showed high accuracy in RCC identification, whereas PR presented poor results deviating significantly from the gold standard. These findings indicate that CBCT imaging contributed effectively to RCC identification. Contrarily, PR was inefficient to identify complex canal configurations. Although CBCT imaging appears to be the most effective imaging examination applicable clinically for identifying RCC, clinicians should always respect the indication criteria and be aware of the limitations of CBCT imaging in displaying the configurations in some cases.
Acknowledgments The authors thank CAPES Foundation (Ministry of Education of Brazil, Brasılia/DF–Brazil) for granting a scholarship to the first author during this research. The authors deny any conflicts of interest related to this study.
References
2. Siqueira JF Jr. Endodontic infections: concepts, paradigms, and perspectives. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 2002;94:281–93. 3. Vertucci FJ. Root canal morphology and its relationship to endodontic procedures. Endod Topics 2005;10:3–29. 4. Yilmaz Z, Tuncel B, Serper A, et al. C-shaped root canal in a maxillary first molar: a case report. Int Endod J 2006;39:162–6. 5. Fan B, Gao Y, Fan W, et al. Identification of a C-shaped canal system in mandibular second molars-part II: the effect of bone image superimposition and intraradicular contrast medium on radiograph interpretation. J Endod 2008;34:160–5. 6. Scarfe WC, Levin MD, Gane D, et al. Use of cone beam computed tomography in endodontics. Int J Dent 2009;2009:634567. 7. Patel S, Durack C, Abella F, et al. Cone beam computed tomography in endodontics a review. Int Endod J 2015;48:3–15. 8. de Pablo OV, Estevez R, Peix Sanchez M, et al. Root anatomy and canal configuration of the permanent mandibular first molar: a systematic review. J Endod 2010;36: 1919–31. 9. Tian YY, Guo B, Zhang R, et al. Root and canal morphology of maxillary first premolars in a Chinese subpopulation evaluated using cone-beam computed tomography. Int Endod J 2012;45:996–1003. 10. Kottoor J, Albuquerque D, Velmurugan N, et al. Root anatomy and root canal configuration of human permanent mandibular premolars: a systematic review. Anat Res Int 2013;2013:254250. 11. Silva EJNL, Nejaim Y, Silva AV, et al. Evaluation of root canal configuration of mandibular molars in a Brazilian population by using cone-beam computed tomography: an in vivo study. J Endod 2013;39:849–52. 12. Paes Da Silva Ramos Fernandes LM, Rice D, Ordinola-Zapata R, et al. Detection of various anatomic patterns of root canals in mandibular incisors using digital periapical radiography, 3 cone-beam computed tomographic scanners, and microcomputed tomographic imaging. J Endod 2014;40:42–5. 13. Neves FS, Freitas DQ, Campos PS, et al. Evaluation of cone-beam computed tomography in the diagnosis of vertical root fractures: the influence of imaging modes and root canal materials. J Endod 2014;40:1530–6. 14. Vertucci FJ. Root canal anatomy of the human permanent teeth. Oral Surg Oral Med Oral Pathol 1984;58:589–99. 15. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Published by: International Biometric Society Stable 1977;33:159–74. 16. Michetti J, Maret D, Mallet JP, et al. Validation of cone beam computed tomography as a tool to explore root canal anatomy. J Endod 2010;36:1187–90. 17. Jaju PP, Jaju SP. Cone-beam computed tomography: time to move from ALARA to ALADA. Imaging Sci Dent 2015;45:263–5. 18. Matherne RP, Angelopoulos C, Kulild JC, et al. Use of cone-beam computed tomography to identify root canal systems in vitro. J Endod 2008;34:87–9. 19. Zhang D, Chen J, Lan G, et al. The root canal morphology in mandibular first premolars: a comparative evaluation of cone-beam computed tomography and microcomputed tomography. Clin Oral Investig 2016 May 16; http://dx.doi.org/10.1007/ s00784-016-1852-x [Epub ahead of print]. 20. Bauman R, Scarfe W, Clark SJ, et al. Ex vivo detection of mesiobuccal canals in maxillary molars using CBCT at four different isotropic voxel dimensions. Int Endod J 2011;44:752–8. 21. Hassan B, Metska ME, Ozok AR, et al. Comparison of five cone beam computed tomography systems for the detection of vertical root fractures. J Endod 2010;36: 126–9. 22. Kamburoglu K, Murat S, Kolsuz E, et al. Comparative assessment of subjective image quality of cross-sectional cone-beam computed tomography scans. J Oral Sci 2011; 53:501–8. 23. Liedke GS, Spin-Neto R, Vizzotto MB, et al. Diagnostic accuracy of cone-beam computed tomography sections with various thicknesses for detecting misfit between the tooth and restoration in metal-restored teeth. Oral Surg Oral Med Oral Pathol Oral Radiol 2015;120:131–7. 24. Th€onissen P, Ermer MA, Schmelzeisen R, et al. Sensitivity and specificity of cone beam computed tomography in thin bony structures in maxillofacial surgery – a clinical trial. J Craniomaxillofac Surg 2015;43:1284–8.
1. Okumura T. Anatomy of the root canals. J Am Dent Assoc 1927;14:632–9.
4
Sousa et al.
JOE — Volume -, Number -, - 2017