Sci—Thur PM: Imaging — 07: Assessment of iterative reconstruction algorithms in four commercial MDCT scanners: a phantom study

Authors

  • Omotayo Azeez,

    1. CancerCare Manitoba, Winnipeg, Manitoba
    2. Department of Radiology, University of Manitoba, Winnipeg, Manitoba
    3. Department of Physics and Astronomy, University of Manitoba, Winnipeg, Manitoba
    Search for more papers by this author
  • Elbakri Idris

    1. CancerCare Manitoba, Winnipeg, Manitoba
    2. Department of Radiology, University of Manitoba, Winnipeg, Manitoba
    3. Department of Physics and Astronomy, University of Manitoba, Winnipeg, Manitoba
    Search for more papers by this author

Abstract

We performed an objective phantom-based image quality evaluation of five commercial iterative reconstruction methods available on four different multi-detector CT (MDCT) scanners at different dose levels as well as the conventional filtered back-projection (FBP) reconstruction. Using the Catphan500 phantom, we evaluated the CT uniformity, CT number accuracy, noise, modulation transfer function (MTF) and noise-power spectrum (NPS). The reconstruction algorithms were evaluated over a CTDIvol range of 0.75 – 18.7 mGy on four major MDCT scanners: GE DiscoveryHD750 (ASIR™ and VEO™); Siemens Somatom Definition AS+ (SAFIRE™); Toshiba Aquilion64 (AIDR3D™); and Philips Ingenuity iCT256 (iDose4™). Images were reconstructed using FBP and the respective iterative reconstructions on all scanners. CT number accuracy and CT uniformity were not affected by the choice of reconstruction method on all scanners. In the dose range of 1.3 – 1.5 mGy, noise reduction compared to FBP using iterative reconstruction was 11% – 51% on GE; 10% – 52% on Siemens; 49% – 62% on Toshiba; and 13% – 44% on Philips scanners. Most algorithms did not affect the MTF, except for VEO™ which produced an increase in the limiting resolution of up to 30%. NPS peak shifted towards lower frequencies NPS amplitude decreased with all iterative algorithms. Compared to FBP, iterative algorithms reduce image noise and increase CNR at the same dose. Noise propagation is Poisson distributed with FBP and ASIR™, while VEO alters the noise dependence with dose. The iterative algorithms available on different scanners achieved different levels of noise reduction and spatial resolution improvements.

Ancillary