A comprehensive evaluation of angular range and separation on image quality, image registration, and imaging dose for cone beam computed tomography in radiotherapy

We aimed to comprehensively evaluate the quality, dose, and registration of cone beam computed tomography (CBCT) images acquired with different angular range (θ) and angular separation (Δθ). Elekta XVI volumetric imaging system was used in this study. CBCT images of a Catphan 503 phantom were reconstructed with projections that were acquired under 17 × 8 different combinations of θ (200° to 360°) and Δθ (0.27° to 3.24°). The evaluation indices were contrast-to-noise ratio (CNR), spatial resolution (modulation transfer function [MTF50%]), uniformity, and registration accuracy. The weighted computed tomography dose index (CTDIw) of CTDI Head dosimetry phantom was measured for estimating radiation exposure. CNR changed little from θ = 200° to 260°, but increased by ≈ 2.0 times from θ = 260° to 360°. CNR also increased by ≈ 2.5 times from Δθ = 2.16° to 0.27°. MTF50% increased with Δθ being decreased, but had no dependence on θ. Image uniformity became better when θ was smaller or larger than 260°, but there was no correlation with Δθ. Registration of the overall body had a maximum error of 0.3 mm. The error of low-contrast objects' registration could reach as much as 1.3 mm. CTDIw showed a linear dependence on θ and 1/Δθ. Although the imaging dose is reduced with θ being decreased or Δθ being increased, the image quality may be degraded. Accuracy of image registration of a rigid body can be achieved; however, low-contrast objects' registration using images with inferior quality will introduce unacceptable errors for clinical use. Further, the benefits and extra dose received by the patients should be balanced, and scan protocols with θ between 210° and 260° or Δθ smaller than 0.54° should not be recommended.