Purpose: The first clinical use of a phantom for unified quality control testing of both cone-beam CT (CBCT) and helical MDCT is reported, with automated analysis and structured reporting of quantitative imaging performance and dose in longitudinal studies spanning 6 clinical systems (3 CBCT + 3 MDCT) for 6 months.
Methods: The phantom (CorgiTM, The Phantom Lab) was used in monthly QC of 3 CBCT systems – Synergy XVI (Elekta), O-arm (Medtronic), and Artis Zee (Siemens) – and 3 MDCT systems – Edge (Siemens), Precision (Canon), and Aquilion One (Canon). Automated software provided one-click analysis and structured reporting of voxel value linearity, contrast-to-noise ratio (CNR), image uniformity, spatial resolution (MTF in axial and non-axial planes), noise, noise-power spectrum (NPS), and cone-beam artifact (signal modulation about two parallel disks). For each system, image quality tests were performed for standard Head (with 13.5 cm diameter phantom) and Body protocols (phantom within a 32 cm annulus).
Results: The results demonstrate a unified system for more rigorous, quantitative, objective QC across a broad spectrum of CBCT and MDCT systems in clinical applications ranging from image-guided radiotherapy, image-guided surgery, interventional radiology to diagnostic radiology and cardiology. Quantitative tests revealed factors not conventionally evaluated and had positive impact on QC of each system – example, quantitative analysis of spatial resolution (MTF, rather than subjective evaluation and line-pair tests that do not probe the limits of spatial resolution on modern CBCT or MDCT systems), the extent to which spatial resolution is isotropic (MTF in axial and non-axial planes), and NPS analysis that was valuable in identifying shortfalls in detector and system geometry calibration.
Conclusion: The phantom, test procedures, and automated analysis of quantitative QC metrics provide a valuable basis for more standardized, rigorous, and quantitative QC in CBCT and with helical MDCT.
Funding Support, Disclosures, and Conflict of Interest: Boone JM & Siewerdsen JH have licensing agreement with The Phantom Lab (Greenwich NY) for the phantom used in the work. Uneri A has a consulting agreement with The Phantom Lab (Greenwich NY) for software development. Mahesh M and Fernandez M do not have any COI related to this study.