Purpose: Cone beam computed tomography (CBCT) has the disadvantage of providing non-quantitative results for bone density determination. The aim of this study is to calibrate CBCT results by using an internal reference (such as muscle) for quantitatively assessing bone density.
Methods: We developed a new method using the relative attenuation ratio between two nearby materials (such as bone and muscle) for systemic error correction in CBCT that depends on the relative object position in the image volume. Phantom calibration was performed to calculate the acquired attenuation ratio in Hounsfield units (HU), comparable to the results from clinical multislice spiral computed tomography (MSCT). In addition, a small animal study with an osteoporotic rat model was evaluated to show the feasibility of this presented method to quantitatively assess bone density using a CBCT system.
Results: The phantom study results showed that the calibration process successfully corrected the systemic inaccuracy from CBCT, and the calibrated HU values agreed with the values measured from MSCT. In the small animal study, the quantitative bone densities assessed from the calibrated CBCT results were consistent with the results from MSCT data.
Conclusion: A practical method to quantitatively estimate attenuation (HU) values for bone tissues from CBCT scans that are comparable to MSCT scans is proposed. The method may improve the quantification ability of CBCT scanning without any additional hardware requirements.