Determination of a cone-beam CT low-dose protocol for root fracture diagnosis in non-endodontically treated anterior maxillary teeth

Abstract

Objectives: The aim of this study was to determine a "low-dose protocol" which provides acceptable diagnostic accuracy for detection of root fractures in unrestored anterior maxillary teeth, using an ex vivo model.

Methods: 48 maxillary anterior teeth, half with horizontal or oblique root fractures, were imaged using CBCT in an anthropomorphic model. Nine X-ray exposure combinations were used, including the manufacturer's standard ("reference") exposure and high-resolution settings ("HiRes"), by varying kV, exposure time, and rotation angle. Measurements of Dose Area Product (DAP) were recorded. Five dental radiologists assessed the scans for root fractures and judged image quality. Parameters of diagnostic accuracy were calculated, including area under the Receiver Operating Characteristic curve (A_z). Objective measures of image quality were made at the same exposure combinations using an image quality phantom.

Results: Although there was a significant linear relationship between DAP and mean A_z, only the lowest DAP exposure combination had a mean A_z significantly different to the reference exposure. There was no significant effect on other diagnostic accuracy parameters when using HiRes compared with the reference exposure. There was a significant positive relationship between DAP and contrast resolution. HiRes did not significantly improve contrast resolution and made a small improvement to spatial resolution.

Conclusions: Scope existed for radiation dose reduction compared with the manufacturer's guidance. There was no improvement in diagnostic accuracy using HiRes settings. A cautious recommendation for this CBCT machine is that it is possible to achieve a dose reduction of about 20% compared with the reference exposure parameters.

Keywords: Cone-beam computed tomography; Diagnosis; Radiation dosage; Radiologic; Technology; Tooth fractures.

Figure 1.

Cross-sectional images of the same root, oriented identically, taken from CBCT scans at each of the nine exposure combinations. There is an oblique fracture line running through the apical and middle thirds of the root. Images at HalfR/70kV, HalfR/80kV, STD/70kV and HiRes/70kV appear “noisy” or “grainy” and it is difficult to see the fracture line clearly except for HiRes70, in which a faint radiolucent line is seen.

Figure 2.

Example cross-sectional image taken from the micro-CT scan of the same tooth as in Figure 1.

Figure 3.

Receiver operating characteristic (ROC) curves for each of the nine exposure combinations, derived from mean data from the five radiologists in the study. For clarity, three graphs are presented, one for each of the operating modes. (a) “half-rotation standard” (HalfR), (b) “full rotation standard” (STD) and (c) “Hi-Resolution” (HiRes), although the reference exposure combination of STD/90kV is included on all.