View the CAT printer-friendly / share this CAT
Title Cone Beam Computed Tomography (CBCT) with Voxels < 0.2 mm3 have Superior Sensitivity to Diagnose Vertical Root Fractures, Despite No Greater Specificity than Periapical Radiography
Clinical Question What is the appropriate voxel size to utilize when performing CBCT imaging to assist in determining the presence of a vertical root fracture (VRF)?
Clinical Bottom Line CBCT scans using voxels < 0.2 mm3 are more predictive of an existing VRF than larger voxel sizes; however, further in vivo studies comparing smaller voxels and radiographic interpretation are warranted.
Best Evidence (you may view more info by clicking on the PubMed ID link)
PubMed ID Author / Year Patient Group Study type
(level of evidence)
#1) 25442497Corbella/20143 in vivo clinical studiesSystematic Review and Meta-analysis
Key resultsThe authors examined the only three clinical studies available at the time of article submission. Three different voxel sizes were examined (0.080 mm3, 0.125 mm3, and 0.2 mm3). The weighted mean scores for sensitivity, specificity, and accuracy for 0.080 mm3 are 1.00, 0.50, 0.90. The weighted mean scores for sensitivity, specificity, and accuracy for 0.125 mm3 are 0.89, 0.94, 0.91. Lastly, the weighted mean scores for sensitivity, specificity, and accuracy for 0.2 mm3 are 0.875, 0.680, 0.790. Conversely, the weighted mean scores for sensitivity, specificity, and accuracy for periapical radiography (PR) are 0.263, 1.00, 0.481, respectively. Receiver operating characteristic (ROC) values were preferable for 0.125-mm3 voxels. Sensitivity was superior for 0.080 mm3, but not significantly different than 0.125 mm3. Specificity was superior for 0.125 mm3, with no specificity (rule-out) difference between CBCT vs. PR.
#2) 25154316Chavda/201421 non-treated and endodontically treated teeth in 20 patientsGold standard-controlled diagnostic study
Key resultsThis clinical study imaged teeth using CBCT with voxels of 0.125 mm3 and 2-dimensional digital PR. Images were interpreted by endodontists and endodontic residents for the presence of VRF. The direct visualization of the presence or absence of vertical fractures after atraumatic extraction of the teeth was considered the “gold standard.” Both PR and CBCT showed poor sensitivity (0.16 and 0.27) and high specificity (0.91 and 0.83) in detecting VRF, with no statistical difference between the two. Receiver operating characteristics (ROC) values were 0.535 and 0.552 for PR and CBCT, respectively. Kappa values for PR and CBCT imaging revealed inter-examiner agreement of 0.496 and 0.532 and intra-examiner agreement of 0.387 and 0.464.
Evidence Search CBCT, vertical root fracture, in vivo, gold standard
Comments on
The Evidence
Validity: For both studies, the gold standard referenced was direct microscopic visualization of the fracture. Moreover, Chavda et al. utilized optical coherence tomography to measure crack width. In Corbella et al., two of the three studies (Metska, Edlund) included only previous endodontically treated teeth, and the 0.2-mm3 study (Metska) used endodontists as examiners while the other two studies utilized oral radiologists. Chavda et al. noted that the "high specificity may also indicate that neither imaging modality is capable of accurately detecting fractures and are in fact producing a disproportionately high number of negative results." This study had poor ROC values, as an ROC area under the curve of 0.5 indicates no one is better at identifying a true positive than flipping a coin. Their kappa values indicate examiners disagreed more with themselves than other examiners. A common day CBCT version of ‘[w]ho’s Reading the Radiograph?’ (PMID: 4501172). Perspective: In vitro studies inaccurately represent clinical x-ray photon attenuation and resultant scatter radiation, which introduces fog and degrades image quality. Therefore, only in vivo studies were reviewed. Spatial resolution refers to how close two objects can be and still be distinguished as two separate objects. In CBCT, the lower limit of resolution is two adjacent voxels. A 0.080-mm3 voxel will require segment separation greater than 0.160 mm (slightly larger than the tip diameter of a #15 K-file) whereas a 0.125-mm3 voxel requires segment separation greater than the tip diameter of a #25 K-file. Smaller voxels can introduce additional noise and artifacts. To overcome this and improve the signal-to-noise ratio, a higher kVp and/or mA could be used, but this will significantly increase the patient’s dose while not necessarily improving diagnostic ability. The concept of ALARA (as low as reasonably achievable) should be kept in mind.
Applicability When CBCT imaging patients suspected of VRF with/or without root canal filling, select voxels < 0.2 mm3 and a localized as possible field of view; oral radiologists present better ROC interpretation values than endodontists.
Specialty/Discipline (Oral Medicine/Pathology/Radiology) (Endodontics) (General Dentistry) (Oral Surgery) (Periodontics) (Prosthodontics) (Restorative Dentistry)
Keywords cone beam computed tomography (CBCT), vertical root fracture (VRF), voxel, in vivo, gold standard
ID# 2779
Date of submission: 11/20/2014spacer
E-mail ADCOCKJ@uthscsa.edu
Author Jeffrey M. Adcock, DDS
Co-author(s) e-mail
Faculty mentor/Co-author S. Thomas Deahl, II, DMD, PhD & Anibal R. Diogenes, DDS, PhD
Faculty mentor/Co-author e-mail DEAHL@uthscsa.edu; DIOGENES@uthscsa.edu
Basic Science Rationale
(Mechanisms that may account for and/or explain the clinical question, i.e. is the answer to the clinical question consistent with basic biological, physical and/or behavioral science principles, laws and research?)
post a rationale
None available
Comments on the CAT
post a comment
None available

Return to Found CATs list