Identification of CBCT Cephalometric Reference Points Under Altered Exposure Settings

Objectives: Introduction: Orthodontics increasingly employs CBCT for cephalometry. Measures to optimize radiation dose while maintaining adequate image quality are essential in this typically young patient population.
Aim: To assess radiation dose to the eye lens, parotid and thyroid gland using various CBCT technique parameters, and to characterize changes in the identification accuracy of typical cephalometric reference points.
Methods: Material and Methods: A human cadaver was imaged in a CBCT machine (Morita Accuitomo 170®) using various acquisition parameters (60, 70, 80, 90 kV; 1, 2.5, 5 mA; 180°, 360° rotation arc). For each acquisition, one dosimeter was positioned on the skin surface overlying each organ of interest (eye lens, parotid gland, thyroid gland). Reconstructed cephalometric images were assessed by 10 orthodontists, with measurements compared to mean error matched by rater and scan settings with respect to a standard cephalometric image (permutation tests).
Results: Results: When decreasing the arc angle from 360° to 180° or reducing the mA from 5 to 1, the dose to the eye lens decreased significantly by 87.5% and 78% respectively. Only for the parotid gland did the absorbed dose significantly decrease approximately linearly with mA: 61%, 50% and 80% for reductions 2.5 to 1 mA, 5 to 2.5 mA and 5 to 1 mA respectively. Regarding cephalometric reference points, a significant increase in measurement variance was shown when using 60 or 70 kV and when decreasing the mA below the default protocol. No statistically significant measurement differences were observed between images taken at 360° or 180° rotation.
Conclusions: Conclusion: For cephalometric analysis, it is recommended to retain the manufacturer’s mA and kV exposure settings (90kV and 5 mA), but to reduce the rotation arc from 360° to 180°, resulting in a 50% dose reduction, without compromising the accuracy of identification of cephalometric reference points.