Image quality evaluation of in-treatment four-dimensional cone-beam computed tomography in volumetric-modulated arc therapy for stereotactic body radiation therapy.

In this study, the image quality of in-treatment four-dimensional cone-beam computed tomography (In-4D-CBCT) obtained with various prescription doses (PDs) were quantitatively evaluated in volumetric-modulated arc therapy (VMAT) for stereotactic body radiation therapy (SBRT) of the lungs and liver. To assess image quality, we used a dynamic thorax phantom and three-dimensional (3D) abdominal phantom; In-4D-CBCT images were acquired with various PDs (from 5 to 12 Gy). The In-4D-CBCT with various PDs were compared with the reference images (pre-4D-CBCT). The image quality was evaluated using the signal-to-noise ratio (SNR), the contrast-to-noise ratio (CNR), and the Dice similarity coefficient (DSC). The fiducial marker positions with various PDs were compared with those of the reference images. For the dynamic thorax phantom, the difference between pre- and In-4D-CBCT in terms of SNR and CNR decreased, as the PD increased from 6 to 12 Gy. The median DSC ranged from 0.7 to 0.74, and showed good similarity. For the 3D abdominal phantom, the difference between pre- and In-4D-CBCT in terms of SNR and CNR decreased as the PD increased from 5 to 6 Gy; conversely, it increased as the PD increased from 7 to 8 Gy. The fiducial marker positions were within 1.0 mm for all PDs. We concluded that the image quality of In-4D-CBCT degraded compared with the reference image; however, it was sufficiently accurate for assessing the intra-fractional tumor position in VMAT for SBRT of the lungs and liver both in terms of the target volume similarity and accuracy of the fiducial marker position.