Room: AAPM ePoster Library
Purpose: In some cases of upper extremity IGRT where the treatment isocenter is laterally displaced from the torso, only a partial-arc CBCT scan is possible. Partial-arc scans have a small (~25 cm) field of view (FOV) which can exclude a large volume of patient tissue, leading to image artifacts and low image quality. In this study we investigated the image quality of CBCT partial-arc protocols as a function of kV and CTDI.
Methods: Eight CBCT protocols (using 100 kV, 125 kV and 140 kV) with CTDIs ranging from 0.32 to 2.4 cGy were studied on a Varian TrueBeam system. Image quality was assessed by scanning an ACR CT phantom with a TomoTherapy "cheese" phantom (30 cm diameter) placed laterally outside of the FOV to simulate the patient body. The ACR phantom images were analysed for uniformity, noise, contrast-to-noise ratio (CNR), Hounsfield unit (HU) accuracy and spatial resolution. The study was performed on six Varian TrueBeam systems to assess machine-to-machine variations.
Results: Image noise was reduced as a function of CTDI, however, there was no obvious increase in CNR with CTDI due to significant out-of-FOV image artifacts. Large HU inaccuracies were observed for all protocols, with the 140 kV scans experiencing the greatest machine-to-machine variations. The limiting spatial resolution increased slightly (5 to 6 lp/cm) with CTDI, while uniformity was unaffected by CTDI or kV. Qualitatively, images acquired with larger CTDIs appeared to be less influenced by streaking artifacts.
Conclusion: In partial-arc CBCT scans with large out-of-FOV artifacts, some image quality improvement is achievable by using a protocol with an increased CTDI. However, the relationship between dose and image quality is not straightforward. In clinical practice, the protocol with the lowest CTDI with acceptable image quality for the task should be used.