Assessment of TrueBeam kV Cone-Beam CT Image Quality and the Clinical Impact of Calibration Drift

Abstract

Background: This study compared the image quality of the TrueBeam (TB) kilo-voltage cone-beam CT (CBCT) system using standard reconstruction (TB-sCBCT) and iterative reconstruction (TB-iCBCT) with that of the Halcyon iterative CBCT (H-iCBCT). Additionally, the clinical impact of calibration drift over a one-year period was investigated for both head and pelvis imaging modes. Methods: Image quality and calibration were assessed using the Catphan®. Image quality evaluation included the analysis of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), and lowcontrast detectability. Calibration stability for head and pelvis CBCT modes was examined by comparing dose-volume histogram (DVH) parameters across ten tongue and ten prostate patient cases. Results: The average SNR was 3.1 for TB-sCBCT, 23.1 for TB-iCBCT, and 6.9 for H-iCBCT, while the corresponding CNR values were 11.2, 53.0, and 28.4, respectively. At a contrast level of 0.3%, the smallest detectable target diameters were 15 mm for TB-sCBCT, 8 mm for TB-iCBCT, and 9 mm for H-iCBCT. Over a one-year period, the Hounsfield Unit (HU) variation in the head CBCT mode exceeded the recommended action threshold of ±50 HU, whereas the pelvis CBCT mode remained within acceptable limits. Conclusion: In the TB system, advanced iterative reconstruction algorithms significantly enhance SNR and CNR by effectively reducing image noise. Halcyon provides improved image quality compared to TB with standard reconstruction; however, it remains inferior to TB-iCBCT. HU deviations were more pronounced in head CBCT than in pelvis mode, although HU variation in the pelvis CBCT mode exhibited greater sensitivity regarding clinical impact.