| dc.description.abstract | The increasing acceptance of CBCT as a standard form of dental diagnostic imaging is based upon its ability to produce multiplanar cross-sectional (3D) images at a level of resolution that, while not as high as conventional radiographs, overcomes anatomical overlap and distortions characteristic of two dimensional (2D) images. CBCT imaging is achieved at relatively low doses compared with conventional CT. Recent promotion of new hybrid devices incorporating 2D and 3D imaging has increased CBCT use in both the primary and specialist dental practice, not just as a pre-operative diagnostic imaging tool but also for post-operative outcome assessment. However, persuasive marketing, has downplayed the increased patient dose, estimated to be between 10–100 times (small field of view [FOV]) that of a standard periapical radiographic exposure, hence significantly increasing radiation detriment. Therefore, it is essential to understand the image quality determinants for CBCT, to fully appreciate the potential for dose optimisation, and to facilitate the establishment of low dose strategies for dental applications. This is particularly relevant for tasks deemed to require higher resolution and doses, such as endodontic imaging. In order to balance radiation risk, the potential benefits of CBCT imaging must be verified, which can only be achieved by analysing the evidence base on the diagnostic efficacy of CBCT relating to particular diagnostic tasks.
The overarching aim was to identify a strategy in which radiation dose can be reduced for a specific diagnostic task, while maintaining images of diagnostic quality, and to demonstrate any potential impact of CBCT imaging on treatment outcome for this task. Initially, the literature on optimisation and diagnostic efficacy relating to CBCT imaging was reviewed. Thereafter, the aim was to identify a dose as low as diagnostically acceptable and a threshold level of image quality for CBCT imaging of root canal anatomy, using the second mesiobuccal (MB2) canal of extracted maxillary first molars (M1M), of varying complexity. Using two CBCT scanners, the aim was also to establish if an objective image quality (IQ) parameter could act as a transferable IQ metric in order to simplify the process of optimisation for a diagnostic task, by using a range of observer experience levels. Subsequently, after an exploration of IQ metrics and IQ determinants in the CBCT imaging chain, the second aim was, using regression analysis, to identify the IQ metric(s) that have a statistically significant relationship with successful achievement of the diagnostic task of root canal identification. A further aim was to identify which of these IQ determinants had a statistically significant relationship with achieving this diagnostic task. A final aim of this thesis was, using 3D-printed M1M teeth, to identify if supplemental CBCT imaging could potentially improve outcome after endodontic access cavity preparation. This was achieved by analysing changes in tooth substance removed, procedural time taken and root canals located, specifically in relation to location of the MB2 canal, compared with the use of conventional (IOPA) and clinical methods alone; thereby justifying the increased dose associated with CBCT imaging.
This is the first study to address optimisation of CBCT imaging of root canal anatomy, and it was shown that a similar threshold dose for this diagnostic task could be identified for both scanners. This optimised dose was achieved using standard protocols instead of high-resolution scans and applied to simple - moderate MB2 canal complexity. Increasing dose to enhance visualisation of more complex canal anatomy was ineffective. The IQ metric, contrast-noise-ratio (CNR) was not demonstrated to be a transferable measure of IQ for this diagnostic task. Further investigation of the relationship between objective and subjective IQ identified CNR as the significant predictor of successfully achieving the diagnostic task of root canal identification when adjusting for other core IQ metrics investigated, using logistic regression. The imaging variables: x-ray tube voltage and current, degree of rotation and native pixel array had a significant effect on successfully identifying root canal anatomy as did experienced observers. The final 3D printed teeth study, showed that additional preoperative CBCT imaging significantly reduced tooth volume removed, procedural time and increased the number of MB2 canals located for standard anatomies and more experienced operators. Within the limitations of the model, these findings may impact patient outcome during root canal treatment but require validation from a prospective clinical trial. | en |
