Since its introduction to dental radiology in 1998, cone-beam computed tomography (CBCT) has been subject to a wealth of scientific reports. The two preeminent queries attempted to be answered are its degree of accuracy in the craniofacial region and the radiological burden of a CBCT examination. The pertinent literature on accuracy can roughly be divided into three groups. Some studies evaluate the influence of device properties and settings, thereby altering voxel-size, varying beam properties (such as kilovoltage, milliampere, field of view, number of projections) or using different post-processing techniques and software. Other studies evaluate the impact different regions of interest have on CBCT accuracy.
A third type of studies assesses the accuracy by comparing CBCT measurements to other imaging methods of the craniofacial region. Several studies were published reporting radiation doses of different CBCT devices or settings. However dose reduction implies also loss of information, and radiation doses should always be estimated in context of diagnostic efficacy to enable a risk-benefit assessment. The aim of this thesis was to perform several investigations considered representative and clinically relevant to establish accuracy and diagnostic efficacy of CBCT, and to determine the radiobiological risk of CBCT examinations in different settings. More specifically, the first aim was to assess CBCT accuracy in the anterior alveolar buccal bone region and to establish the impact of different voxel sizes on accuracy. The second goal was to compare the accuracy of CBCT to multidetector computed tomography (MDCT) with special emphasis on the susceptibility to metal artefacts. A further objective was the direct comparison of accuracy of linear measurements of mandibular ramus height and the condylar process based on data of 3D (CBCT, MDCT and magnetic resonance imaging (MRI)) and 2D imaging. For this, the best-suited MRI sequence for assessment of the condylar process had to be established, using micro-CT as reference. The purpose of the last investigation on diagnostic efficacy was to validate the assessment of fusions and osteoarthritic changes in CBCT and MDCT. To evaluate the radiobiological burden of CBCT, dose analysis at different settings, with and without exposure of the cervical spine was performed with the aim to report a risk-benefit recommendation.
Unmitigated embalmed cadaver heads were used to obtain anatomical records which served as reference and corresponding radiological data of 2D imaging and 3D imaging (CBCT, MDCT, MRI, micro-CT) were generated. The results indicate that voxel size affects precision of CBCT measurements and that alveolar bone of 1mm thickness might be missed completely, even with voxel size of 0.125mm. Compared to MDCT, CBCT appears to be less susceptible to metal artefacts and is slightly more reliable for linear intraoral measurements. CBCT, MDCT and MRI are almost equal for measurements of mandibular ramus height and the condylar process and more reliable than 2D imaging. MRI is also well suited for cortical bone thickness measurement of the mandibular condyle and for evaluation of osteoarthritic changes in the condyle.
Thus, although high reproducibility and precision of CBCT are shown for measurements of ramus height and the condylar process, MRI is recommended since it is not only an equal alternative to CBCT and MDCT, but also circumvents the issue of ionizing radiation. CBCT data screened by oral radiologists are as reliable as MDCT data viewed by general radiologists to exclude fusions in the cervical spine, but general radiologists appraising MDCT perform better in the assessment of osteoarthritis. The results of the radiation dose evaluation demonstrate that reducing the field of view or the application of a thyroid shield are both efficient methods to reduce the effective dose and must therefore be implemented. In conclusion, the clinician's choice over which CBCT setting to use should depend on the intended diagnostic purpose of the scan and on the region of interest. Alternative imaging methods should always be considered and available methods must be implemented to reduce exposure of vulnerable craniofacial tissues.