CT Number and Geometric Measurement Consistency of Dual-Source Vs. Single-Source Spectral Imaging on a Photon-Counting CT: A Phantom Study
Abstract
Purpose
Dual‑source photon‑counting CT (DS‑PCCT) enables fast FLASH acquisition combined with spectral imaging—capabilities not available in single‑source PCCT (SS‑PCCT). This study evaluates whether DS‑PCCT maintains quantitative accuracy by comparing CT number and geometric measurement consistency with SS‑PCCT across multiple phantom sizes and scan parameters.
Methods
Three cylindrical phantoms (20×20 cm, 33×26 cm, 40×30 cm) were scanned using a photon‑counting CT system (NAEOTOM Alpha, Siemens). Seven large rods as inserts (28 mm diameter, iodine 0.5–20 mg/mL, solid‑water, lung‑equivalent) were analyzed for CT numbers, while three small iodine rod inserts (2, 5, 10 mm) were used for geometric measurements. Scans followed a chest CTA protocol at 120 kV with IQ level of 90 and CTDIvol of 1.2, 3.2, and 5.2 mGy, respectively for each phantom. SS‑PCCT used pitches of 0.8 and 1.5, and DS‑PCCT used pitches of 1.5, 2.0, 2.5 and 3.2. Images were reconstructed with 2 mm slices, Bv40f kernel, iterative level 3, and spectral image types (55 keV virtual monoenergetic, virtual noncontrast, iodine maps). In each image, ROIs were placed in each large rod, and small‑rod areas were quantified using adaptive thresholding.
Results
Across all rod‑image‑type combinations, linear mixed‑effects models showed no statistically significant (p>0.05) differences between DS‑ and SS‑PCCT for CT numbers or geometric measurements in 26 of 27 comparisons. Interactions between scan mode (DS vs. SS) and pitch were also not significant in most comparisons. Pitch had no statistically significant effect on area consistency and minor effects on some CT number measurements, with pitch‑related slopes generally <5 HU.
Conclusion
DS‑PCCT demonstrated consistency in CT number accuracy and geometric performance compared with SS‑PCCT, supporting its reliability for routine clinical use.