Table 2 Characteristics of the included studies
Studies | Population/ ethnic group | Sample size and (its estimation method) | Diagnostic device specifications and technique | Evaluation process | Ethical approval | |
|---|---|---|---|---|---|---|
1. | Mannocci et al. [47] | Not reported | 20 (Not reported) | 100 kV to achieve a focal spot of 6 µm. The sample was positioned between the source and the detector to achieve X4 magnification. Two-dimensional individual images with a pixel size of 12.5 µm, and a slice thickness of 25.0 µm were obtained | Axial sections of 3D images | Not reported |
2. | Iwaka et al. [48] | Japanese | 30 (Not reported) | 45 kV, 100 μA and slice width, 16.5 μm. Prior to imaging, each tooth was set such that the sample stage was orthogonal to the tooth axis. (PHOTOSHOP 6.0, Adobe inc., Sanjose, CA, USA) and (VOXBLAST, Vaytec Inc., Los Angeles, CA, USA | 2D & 3D | Patients' agreements |
3. | Gu et al. [49] | Chinese | 36 (Not reported) | Voxel sizes of 15 × 15 × 15 mm, and the cross-section was perpendicular to the long axis of the root | 2D | Not reported |
4. | Gu et al.[12] | Chinese | 122, then 25 2-rooted were selected (Not reported) | Each specimen was scanned along the teeth axis with voxel sizes of 21 mm × 21 mm × 21 mm. The resulting data were then processed by software Microview 2.1.2 (GE HealthCare, London, Ontario, Canada) | 2D &3D | Not reported |
5. | Gu et al. [50] | Chinese | Initially 122, then 25 were selected (Not reported) | Each specimen was scanned along the teeth axis with voxel sizes of 21 × 21 × 21 mm. The data sets (DICOM format) were transferred to Mimics 10.01 (Materialise, Leuven, Belgium) software | 3D | Not reported |
6. | Fan et al. [51] | Chinese | 70 (not reported) | 37-mm intervals, with the scanning being perpendicular to the long axis of the teeth. (3D-Doctor; Able Software Corp, Lexington, MA) was used for 3D reconstruction and image analysis | 3D | Not reported |
7. | Harris et al. [52] | Minneapolis, MN, area | 22 (Not reported) | A series of tomographic images (at approximately 935 × 1001 × 1437 voxels, effective resolution 11.41 × 12.21 × 17.53 mm) were obtained for each specimen. CT Pro software (Nikon Metrology, Leuven, Belgium) was used to reconstruct a 3-dimensional image of each tooth. VG Studio MAX 2.1 software (Volume Graphics GmbH, Heidelberg, Germany) was then used for viewing and measuring | 2D &3D | Not reported |
8. | Filpo-Perez et al.[5] | Brazilian population | 100 (Not reported) | 50 kV, 800 mA, a rotation step of 0.8, 360of rotation, and an isotropic resolution of 19.6 mm. (NRecon v.1.6.9, Bruker-microCT) was used for reconstruction. CTVol v.2.2.1 and Data Viewer v.1.5 software (Bruker-microCT) were used for visualization and qualitative evaluation | 2D & 3D | Ethics committee approval |
9. | Lamia & McDonald [53] | Not reported | 114 (Not reported) | 18-µm voxel size, medium resolution, 90 kV, 88-µA intensity, 0.5-mm aluminium filter, integration time of 500 ms, and 30-µm slices. scan times were approximately 3.5 h per sample | 3D | Not reported |
10. | Lee et al. [54] | Korean | 37 (Not reported) | Voxel size = 31.8 µm3. 3D modelling software V-works 4.0 (Cybermed Inc., Seoul, Republic of Korea) was used for viewing and analysis | 2D & 3D | Institutional Review Board approval |
11. | Gu et al. [55] | Chinese | 45 (Not reported) | Each specimen was scanned along the tooth axis with voxel size of 15 mm or 21 mm by using a micro-CT scanner (Inveon; Siemens Medical Solutions, Knoxville, TN). The data sets were transferred to the Mimics 15.01 (Materialise, Leuven, Belgium) software in DICOM format, and Mimics was used to read and reconstruct the data into 3D images | 3D | Medical Ethics Committee |
12. | Versiani et al. [22] | Brazilian | 25 mandibular first molars with MMCs (Not reported) | 100 kV, 100 µA and an isotropic voxel size of 9.9 µm. The scanning procedure was performed through 180º rotation around the vertical axis, with a rotation step of 0.4º, using a 0.5-mm-thick aluminium filter. (NRecon v.1.6.9; Bruker-microCT) and (CTAn v.1.14.4; Bruker-microCT) software were used for reconstruction and surface representations of the internal anatomy, respectively | 2D &3D | Research Ethics Committee |
13. | Versiani et al. [56] | Brazilian and Turkish populations | 258 Brazilian (136) and Turkish (122), then 48 teeth with MMCs were assessed (Not reported) | 100 kV, 100 mA and an isotropic voxel size of 9.9 mm. Scanning procedure was performed through 180˚ rotation around the vertical axis, with a rotation step of 0.4, using a 0.5-mm-thick aluminium filter. (NRecon v.1.6.9; Bruker-microCT) and (CTAn v.1.14.4; Bruker-microCT) software were used for reconstruction and surface representations of the internal anatomy, respectively | Cross and coronal sections of the 3D images | Research Ethics Committee |
14. | Wolf et al. [57] | Egyptian | 118 (Not reported) | Isotropic resolution of 20 mm, 70 kV and 114 mA, resulting in 800–1200 slices per tooth. (VGStudio Max 2.2; Volume graphics, Heidelberg, Germany) was used for reconstructing and viewing the images | 3D | Not reported |
15. | Gu et al. [58] | Chinese | 25 (Not reported) | Voxel size of 15 or 21 μm. Mimics 15.01 (Materialise, Leuven, Belgium) software was used for reconstructing and viewing the images | 3D | Medical Ethics Committee |
16. | Keles & Keskin [59] | Turkish | Initially 269 then only 83 roots with Vertucci type II (Not reported) | 100 kV and 100 mA. Slices presenting 2000 × 1330 pixel resolution with 10 mm pixel size were obtained from each root by using an 11 MP camera. Scanning was performed at 180 rotations around the vertical axis with a camera exposure time of 1400 ms and a rotation step of 0.4. NRecon software (v. 1.6.4; Bruker-microCT) and CTAn software (v.1.13; Bruker-microCT) were used for the reconstruction and measuring, respectively. Beam-hardening correction of 45%, smoothing of 2, and an attenuation coefficient range of 0–0.06 | 2D & 3D | Ethical board |
17. | Keles & Keskin [26] | Not reported | 85 mandibular molar teeth with MMCs | 100 kV and 100 mA. Slices presenting 2000 X 1330 pixel resolution with 10-mm pixel size were obtained from each root using an 11-megapixel camera. Scanning was performed with 180˚ rotations around the vertical axis with a camera exposure time of 1400 ms and a rotation step of 0.4˚. Data were reconstructed using NRecon software (v. 1.6.4, Bruker-microCT) with a beam-hardening correction of 45%, smoothing of 2, and an attenuation coefficient range of 0–0.06. CTAn and Data Viewer (v.1.5, Bruker microCT) software were used to present the root canal configuration of each root | 2D | Not reported |
18. | Moe et al. [60] | Myanmar | Initially, 181, then 75 were selected (Not reported) | 10-μm isotropic resolution, 125 μA, 80 kV, 1-mm aluminium filter, and 0.4° rotation step with 180° rotation. NRecon software v1.6.1 (Bruker MicroCT) and CTAn software v1.14.4 (Bruker MicroCT) were used for reconstruction. CTvol software v2.2.3 (Bruker MicroCT) was used for visualizing the images | 2D & 3D | Institutional Review Board |
19. | Wolf et al. [61] | Not reported | 118 (Not reported) | 70 kV and 114 μA, resulting in 800–1200 slices per tooth at an isotropic resolution of 20 μm. (VGStudio Max2.2; Volume-graphics, Heidelberg, Germany) was used to be able to differentiate the tooth structures | Axial and coronal sections of 3D images | Not reported |
20. detecti 21. | Keles & Keskin [62] | Turkish | Mesial roots of 269 teeth, then only 109 selected (Not reported) | 100 kV and 100 mA. Slices presenting 2000 × 1330 pixel resolution with 10-mm pixel size were obtained from each root using an 11-megapixel camera. NRecon software (v. 1.6.4, Bruker-microCT) was used for reconstructing the images with a beam hardening correction of 45%, smoothing of 2, and an attenuation coefficient range of 0 to 0.06. CTAn and DataViewer (v.1.5, Bruker microCT) software were used to reveal the root canal configuration of each root | 2D & 3D | Not reported |
22. | Keles & Keskin [63] | Not reported | 269 mesial roots then only 40 (Not reported) | 100 kV and 100 mA. Slices presenting 2000 × 1330 pixel resolution with 10 mm pixel size were obtained from each root by using an 11 MP camera. Scanning was performed at 180 rotations around the vertical axis with a camera exposure time of 1400 ms and a rotation step of 0.4˚. NRecon software (version 1.6.4, Bruker-microCT) and CTAn software (version 1. 13, Bruker-microCT) were used for image reconstruction and analysis, respectively | 2D & 3D | Ethical board |
23. | Theye et al. [64] | Skulls from South African | 24 (Not reported) | 100 kV voltage, 100 mA current, and 2.00 s exposition time per projection, with an isotropic voxel size ranging from 40 to 48 mm. Nikon CT Pro (Nikon Metrology) and VG Studio MAX-3.0 (Heidelberg, Germany) software were used for reconstructing the images and visualization, respectively | 2D &3D | Research Ethics Committee |
24. | Tomaszewska et al. [4] | Not reported | 108 (Not reported) | Spatial resolution 13.68 µm per pixel CTVox, CTAn alyser and CTVol (SkyScan®) applications were used for reconstruction and visualisation | 3D | Bioethical Commission |
25. | Marceliano-Alves et al [7] | Brazilian | 140 (Not reported) | 50 kV, 120 mA, with a rotation step of 0.8, 360° around the vertical axis, and 12.1 µm pixel size, using a 1-mm-thick aluminium filter. NRecon software (v 1.6.1.0; Bruker, Kontich, Belgium). Reconstruction parameters included a 50% beam hardening correction, ring artefact correction of 5 and smoothing of 5. (CTAn v.1.14.4, Bruker-microCT) were used for reconstruction and measuring, respectively | 2D & 3D | Ethical committee |
26. | Arfianti et al. [65] | Not reported | 19 (Not reported) | Resolution, 50 µm (medium); voltage, 130 kV; current, 60 µA; rotation angle, 240°; and time exposure, 295 ms. NRecon and NRecon Server software were used for reconstruction and analysis. DataViewer and Fiji ImageJ software were used for visualising and measuring the parameters, respectively | 2D | Not reported |
27. | Asijavičienė et al. [66] | Not reported | 60 (Not reported) | 110 kV, 50 mA, 1-mm aluminium filter, 180° rotation around the vertical axis with rotation step of 0.18 and an isotropic resolution of 22.8 μm. (NRecon v.1.6.9, Bruker-microCT) and CTVol 1.10.1.0 software (Bruker-microCT) were used for reconstruction and volumetric visualisation, respectively | 2D &3D | Local ethical committee |
28. | Keles et al. [24] | Turkish | Initially 250 then only 30 for bifid and 30 for non-bifid, (by using [G*Power 3.1]) | 100 µA, 100 kV, 180° rotation with a step of 0.4°, frame average of 3 and 1,400 ms of exposure duration. Pixel size of 10 µm. Data were reconstructed (NRecon v. 1.7.4.2 software; Bruker-microCT) with ring artefact (5), beam-hardening (45%), and smoothing (2) corrections using an attenuation coefficient ranging from 0 to 0.06. CTAn v.1.18.8 software (Bruker-microCT) and CTVol v. 2.3.2.0 software (Bruker microCT) were used for 3D reconstruction and qualitative analysis | 2D & 3D | Ethics committee |
29. | Mazzi-Chaves et al. [67] | Brazilian | 50 (Not reported) | Voxel size of 26.70 μm. The scanning parameters used were 50 kV, 800 μA, 180° rotation around the vertical axis, rotation step of 1°, and a 0.5 mm-thick aluminium filter, rendering a scan time of 25 min, approximately. NRecon v.1.7.1.0 software (Bruker-microCT, Kontich, Belgium) ring artifact reduction of 5, beam hardening correction of 40%, smoothing of 3, and an attenuation coefficient between 0.001 and 0.15. DataViewer v.1.5.4.0 software and (Bruker-microCT, Kontich, Belgium) were used for reconstruction and measuring, respectively. CTAn v.1.17.7.2 + software (Bruker microCT, Kontich, Belgium) for generating 3D models | 3D | Research Ethics Committee |
30. | De-Deus et al. [11] | Brazilian subpopulation | 120 ([G*Power 3.1] software) | 14.25 μm (pixel size), 70 kV, 114 mA, 180° rotation around the vertical axis, rotation step of 0.7°, camera exposure time of 250 ms, frame average of 4, using a 1-mm-thick aluminium filter. (NRecon v. 1.7.1.6; Bruker-microCT) was used for reconstruction with beam hardening (35 to 45%), ring artefact correction (3 to 5), and contrast limits (0 to 0.05). DataViewer v.1.5.6 software (Bruker-microCT) was used for qualitative and quantitative analysis | 2D & 3D | Local ethical committee |
31. | Fu et al. [36] | Chinese | 136 (Not reported) | 90 kV/88 mA with an isotropic voxel size of 30 mm. Scanning was performed by 500 projections per 180, camera exposure time of 500 ms | 3D | Ethics committee |
NO | Studies | Type of study | Type of analysis | Classification used | Calibration and assessment reliability | |
1. | Mannocci et al. [47] | Retrospective | Quantitative | N/A | Two examiners | |
2. | Iwaka et al. [48] | Prospective | Quantitative | N/A | Not reported | |
3. | Gu et al. [49] | Prospective | Quantitative and qualitative | Weller system for isthmus classification | Agreement had been achieved by three observers | |
4. | Gu et al.[12] | Prospective | Quantitative and qualitative | Vertucci system for canal configuration | Not reported | |
5. | Gu et al. [50] | Prospective | Quantitative | N/A | 3 times repeated measurements | |
6. | Fan et al. [51] | Retrospective | Quantitative and qualitative | Fan classification for isthmus | Not reported | |
7. | Harris et al. [52] | Retrospective | Quantitative | Vertucci | Not reported | |
8. | Filpo-Perez et al.[5] | Retrospective | Quantitative and qualitative | Vertucci system for canal configuration | Not reported | |
9. | Lamia & McDonald [53] | Retrospective | Quantitative | N/A | Two examiners | |
10. | Lee et al. [54] | Retrospective | Quantitative | N/A | Not reported | |
11. | Gu et al. [55] | Prospective | Quantitative | N/A | Intra- and inter-observer agreement was estimated on 8 specimen (1 specimen for each root form), and each specimen was measured twice. Intraclass and one-way random effects model were calculated. The interobserver agreement in RSA was higher in both examiners, with an ICC of 0.999 (95% CI: 0.993, 1.000) (p = 0.000) in examiner 1 (Gu Y) compared to correlation coefficients (ICC) based on 0.999 (95% CI: 0.994, 1.000) (p = 0.000) in examiner 2 (Zhu Q). The ICC for inter-observer agreement was 0.994 (95% CI: 0.973, 0.999) (p = 0.000) | |
12. | Versiani et al. [22] | Retrospective | Quantitative | N/A | Not reported | |
13. | Versiani et al. [56] | Retrospective | Quantitative and qualitative | Pomeranz system for middle mesial canal | Not reported | |
14. | Wolf et al. [57] | Retrospective | Quantitative and qualitative | Four-digit system for canal configuration | Not reported | |
15. | Gu et al. [58] | Retrospective | Quantitative | N/A | Not reported | |
16. | Keles & Keskin [59] | Retrospective | Quantitative | Vertucci system for canal configuration | Not reported | |
17. | Keles & Keskin [26] | Retrospective | Quantitative and qualitative | Pomeranz system for middle mesial canal | Not reported | |
18. | Moe et al. [60] | Retrospective | Quantitative and qualitative | Weller for isthmus classification | Not reported | |
19. | Wolf et al. [61] | Retrospective | Quantitative and qualitative | N/A | Not reported | |
20. | Keles & Keskin [62] | Retrospective | Quantitative | N/A | Not reported | |
21. | Keles & Keskin [63] | Retrospective | Quantitative and qualitative | N/A | Not reported | |
22. | Theye et al. [64] | Retrospective | Quantitative | N/A | Not reported | |
23. | Tomaszewska et al. [4] | Prospective | Quantitative | Vertucci for canal configuration | Two researchers did the measurements, and the 3rd one averaged them | |
24. | Marceliano-Alves et al.[7] | Retrospective | Quantitative and qualitative | Vertucci for canal configuration. Hsu & Kim for isthmus | Not reported | |
25. | Arfianti et al. [65] | Retrospective | Qualitative | N/A | Not reported | |
26. | Asijavičienė et al. [66] | Retrospective | Quantitative and qualitative | Fan classification for isthmus | Not reported | |
27. | Keles et al. [24] | Retrospective | Quantitative and qualitative | N/A | Not reported | |
28. | Mazzi-Chaves et al. [67] | Retrospective | Quantitative | Pucci & Reig and AAE for RCC | Not reported | |
29. | De-Deus et al. [11] | Retrospective | Quantitative and qualitative | N/A | Not reported | |
30. | Fu et al. [36] | Retrospective | Qualitative | N/A | Not reported | |