eduzhai > Applied Sciences > Engineering >

X-ray Monochromatic Imaging from Single-spectrum CT via Machine Learning

  • Save

... pages left unread,continue reading

Document pages: 12 pages

Abstract: In clinical CT system, the x-ray tube emits polychromatic x-rays, and thex-ray detectors operate in the current-integrating mode. This physical processis accurately described by an energy-dependent non-linear integral equation.However, the non-linear model is not invertible with a computationallyefficient solution, and is often approximated as a linear integral model in theform of the Radon transform. Such approximation basically ignoresenergy-dependent information and would generate beam hardening artifacts.Dual-energy CT (DECT) scans one object using two different x-ray energy spectrafor the acquisition of two spectrally distinct projection datasets to improveimaging performance. Thus, DECT can reconstruct energy and material-selectiveimages, realizing monochromatic imaging and material decomposition.Nevertheless, DECT would increase radiation dose, system complexity, andequipment cost relative to single-spectrum CT. In this paper, amachine-learning-based CT reconstruction method is proposed to performmonochromatic image reconstruction using a single-spectrum CT scanner.Specifically, a residual neural network (ResNet) model is adapted to map a CTimage to a monochromatic counterpart at a pre-specified energy level. ThisResNet is trained on clinical dual-energy data, showing an excellentconvergence to a minimal loss. The trained network produces high-qualitymonochromatic images on testing data, with a relative error of less than 0.2 .This work has great potential in clinical DECT applications such as tissuecharacterization, beam hardening correction and proton therapy planning.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×