As part of a CBCT Quality Assurance Plan, a set of CBCT Quality Control Test Tools are used for the evaluation of the imaging performance characteristics of a modality, usually comprising a QC phantom and a special software program.
The QC phantom
Preliminary tests on the NewTom 3G CBCT unit showed that using a phantom designed for Quality Assurance on medical CT equipment results in images with worse low-contrast resolution than the medical CT scan. Furthermore, discrimination between objects with different density was not always successful. It is speculated that this is due to the fact that NewTom 3G - and possibly all other dental CBCT units – are optimized for imaging of hard tissues. This is also related to the low dose delivered compared with medical CT. Therefore, the development of a specifically designed phantom, with a size and densities resembling those of dental interest is necessary. A variety of test objects shall be included in the phantom body (as inserts) for the testing of the imaging performance characteristics.
The QC software
Software tools are developed for the interpretation of the phantom imaging results and the evaluation of image quality. Simple QC measurements can be automatically performed if mathematical models of the test objects are registered with real images. It is under investigation whether more complex QC measurements, such as Line Spread Function, Modulation Transfer Function and others can be performed automatically, using specific test object registration modules.
Uniformity and Image noise: Image noise is calculated for a specific uniform area of a phantom at certain dose levels. It is expressed in terms of SD of gray levels within ROIs selected in test objects with uniform densities. Several ROIs are selected at different positions to check if the noise is similar throughout the image and also if the mean level of gray in each area is the same.
High contrast - Spatial resolution (in terms of MTF and lp/cm). A test object with a thin wire placed along the rotation axis is used to calculate the Point Spread Function which leads to the calculation of MTF over the spatial frequencies spectrum in cycles/cm. A test object with bar patterns is used to test the limiting resolution (maximum high contrast resolution) in lp/mm.
Low contrast resolution: Since dental CBCT units are mainly optimized for imaging of hard tissues, the reference background level should be density of cancellous bone instead of water as used on medical CT testing, where low contrast resolution should be determined for several ROI sizes so that a contrast-detail diagram for the low contrast part can be produced. Additionally, in the cancellous bone density background, several test objects with densities of clinical importance may be inserted, like that of cortical bone, fibrous tissue, enamel, dentine, water and air. The ability of a system to differentiate among these densities at certain image noise levels (i) and spatial resolution (ii), would give a good measurement on its efficiency to produce images of high diagnostic value in the maxillofacial region for certain clinical applications (trauma, lesions, implant placement planning, TMJ examination).
Geometric performance: Test objects with patterns should be inserted to facilitate the testing of the geometric performance of the CBCT units. These tests reveal the efficiency of the Cone Beam correction algorithms that rectify geometric errors closer to the circumference of the imaging field in units with large apertures. This is essential for the testing of measurements accuracy in the reconstructed images in all planes. Both integral and differential linearity in X, Y and Z directions should be measured. This can be done using a phantom with small beads inserted with a spacing of 2-3 mm in all three directions.
Artefacts: Combinations of objects that induce artefacts can be used to evaluate the tolerance of the system to the induction of these artefacts.