IADR Abstract Archives
Correcting Motion Artefacts in CBCT and Impact on Image Quality
Objectives: The objective of this study was to assess the impact on image quality of head motion artefacts and their subsequent correction in CBCT images for simulated mandibular implant planning.
Methods: A partially dentate human skull was incorporated into a robot simulating four types of head movement: antero-posterior translation, nodding/swallowing, lateral rotation, and tremor; at three distances (0.75, 1.5, and 3mm). Three CBCT-units examined the skull during the twelve head movements (one per examination) and during no-motion: Cranex 3Dx (CRA, Soredex Oy, Finland), Orthophos SL 3D (ORT, Sirona Dental Systems GmbH, Germany), and X1 (3Shape, Denmark), the latter with (X1wi) and without (X1wo) subsequent motion artefact correction, i.e. four imaging modalities. In total 52 examinations were performed. Three observers, blinded to movement and imaging modality, scored image quality defined as the presence of 1)stripe artefacts (yes/no), 2)unsharpness (yes/no), and 3)unable to make a report on implant planning (in case of severe artefacts/unsharpness) (yes/no). Kappa statistics assessed interobserver agreement. Chi-square-test analyzed whether observers’ ability to report differed between the four imaging modalities.
Results: Interobserver agreement for ability to report was almost perfect (average 0.96). Observers were unable to report in 17.9% (range 17.3-19.2 among observers). On average, 61.5% (range 42.3-76.9%) presented stripe artefacts and 75.9% (range 71.2-78.8) presented unsharpness. For CRA, ORT, and X1wo, artefacts were present for all movement types and distances; for X1wi, artefacts were present only for tremor. Observers were able to report on all examinations for X1wi. For CRA, ORT, and X1wo observers were unable to report on examinations with tremor. For ORT, one observer could not report on examinations with 3-mm antero-posterior translation. Statistically significant difference was thus seen in the ability to report between X1wi and the other three modalities (p=0.001).
Conclusions: Correcting motion artefacts significantly enhanced CBCT image quality and observers’ ability to report.
Methods: A partially dentate human skull was incorporated into a robot simulating four types of head movement: antero-posterior translation, nodding/swallowing, lateral rotation, and tremor; at three distances (0.75, 1.5, and 3mm). Three CBCT-units examined the skull during the twelve head movements (one per examination) and during no-motion: Cranex 3Dx (CRA, Soredex Oy, Finland), Orthophos SL 3D (ORT, Sirona Dental Systems GmbH, Germany), and X1 (3Shape, Denmark), the latter with (X1wi) and without (X1wo) subsequent motion artefact correction, i.e. four imaging modalities. In total 52 examinations were performed. Three observers, blinded to movement and imaging modality, scored image quality defined as the presence of 1)stripe artefacts (yes/no), 2)unsharpness (yes/no), and 3)unable to make a report on implant planning (in case of severe artefacts/unsharpness) (yes/no). Kappa statistics assessed interobserver agreement. Chi-square-test analyzed whether observers’ ability to report differed between the four imaging modalities.
Results: Interobserver agreement for ability to report was almost perfect (average 0.96). Observers were unable to report in 17.9% (range 17.3-19.2 among observers). On average, 61.5% (range 42.3-76.9%) presented stripe artefacts and 75.9% (range 71.2-78.8) presented unsharpness. For CRA, ORT, and X1wo, artefacts were present for all movement types and distances; for X1wi, artefacts were present only for tremor. Observers were able to report on all examinations for X1wi. For CRA, ORT, and X1wo observers were unable to report on examinations with tremor. For ORT, one observer could not report on examinations with 3-mm antero-posterior translation. Statistically significant difference was thus seen in the ability to report between X1wi and the other three modalities (p=0.001).
Conclusions: Correcting motion artefacts significantly enhanced CBCT image quality and observers’ ability to report.