Residual rotational error of X-ray volume imaging system and its dosimetric effect in spinal tumor radiation therapy.

This study was designed to investigate the residual rotational error (RRE) of an X-ray volume imaging (XVI) system and evaluate the dosimetric impact of this error on spinal tumor radiation therapy. Various rotational displacements (set rotations) were applied to an anthropomorphic phantom placed on a HexaPod evo RT CouchTop (HexaPod couch). To detect these set rotations, a series of cone-beam computed tomography (CBCT) scans of the phantom were acquired and registered to the planning CT in the XVI system. The RRE of the XVI system was evaluated by comparing the difference between the set rotations and the registration results from the XVI. The error-introduced plans (by applying the RRE to the copies of the reference plan) were generated in the treatment planning system. The dose distribution was compared between the reference plan and the error-introduced plans to assess the dosimetric impact of RRE. The root-mean-square (RMS) of RREs were 0.31°, 0.35°, and 0.25° in the X (pitch), Y (roll), and Z (yaw) direction, respectively. For the reference plan versus the error-introduced plans, the PTV volumes receiving the prescribed dose (V 100) were 95.1 % versus 94.8-95.7 %; the conformity indices of the PTV were 1.17 versus 1.16-1.19; the minimum dose to 1 cc of volume (D1 cc) of spinal cord were 43.73 Gy versus 43.71-43.89 Gy; the left kidney volumes receiving 15 Gy (V 15) were 29.7 % versus 29.2-30.7 %; and the V 15 values of the right kidney were 26.1 % versus 24.6-27.5 %. Relative to the reference plan, the dose difference of error-introduced plans exceeded 3 % in kidney V 15. In conclusion, the XVI system can accurately detect the rotational displacement. However, large dose deviations were introduced by RREs when organs at risk were away from the iso-center even for small RREs.