The Image Quality of Ion Computed Tomography at Clinical Imaging Dose Levels
DC Hansen1*, TS Sorensen1, J Seco2, (1) Aarhus University, Aarhus, Denmark, (2) Mass General Hospital; Harvard Medical, Boston , MAWE-G-141-5 Wednesday 4:30PM - 6:00PM Room: 141
To investigate the resolution and image quality that can be obtained from ion computed tomography (CT) at clinical dose levels (~10mSv) and number of scanning angles.
A 20cm cylindrial phantom was designed in the Monte Carlo code Geant4 with the explicit purpose of obtaining the edge-spread function (ESF) and contrast levels. An ion-CT scan of the phantom was simulated for 230MeV, 330MeV protons and 430MeV/u carbon-ions. These were chosen as the represent the limits of currently available technology. Each simulation was repeated for a varying total number of scanning angles, ranging from 20-360. The target dose was set to 10mSv, similar to what is given in a standard x-ray CT scan. The conversion from physical to biological dose was done following the publications of ICRP. The CT images were reconstructed using cubic spline curves for the proton trajectories.
In addition to the ion-CT images, the 230MeV proton images were also reconstructed using prior knowledge obtained from a simulated CT of the phantom.
Resolution was defined as the σ of the Gaussian ESF.
In general the combined proton and x-ray scans yielded the highest resolution (0.4mm) followed by 330MeV protons (0.8mm) and 230 MeV protons (1.0mm), with carbon ions giving the lowest resolution (1.1mm).
When compared with the ground truth, the combined x-ray and proton image yielded the best result, followed by 230MeV protons, 330MeV protons and carbon ions yielding the worst result.
Both 230MeV and 330MeV protons have distinct advantages for ion-CT, and for clinical scans the energy should be decided on a case by case basis. Carbon-ions, while in theory superior, cannot be recommended for patient scans, due to the high dose level. The addition x-ray CT prior knowledge into the reconstruction of ion-CT yields a better image than either ion-CT or x-ray CT alone.
Funding Support, Disclosures, and Conflict of Interest: This work was supported by the Danish Center for Interventional Research in Radiation Oncology (CIRRO).