Click here to


Are you sure ?

Yes, do it No, cancel

X-Ray Tube Monte Carlo Model Based On GEANT4 to Optimize Excitation Energy Spectrum and Source Collimator Designfor X-Ray Fluorescence Imaging

R Schmidt*, J Shi, J Ford, N Dogan, A Pollack, University of Miami Miller School of Medicine, Miami, FL


(Sunday, 7/12/2020)   [Eastern Time (GMT-4)]

Room: AAPM ePoster Library

Purpose: To optimize the excitation energy spectrum for K-shell x-ray fluorescence (XRF) imaging of gold nanoparticles (GNPs), we developed a computational model (based on Geant4) of a benchtop x-ray tube to optimize the design of excitation filter and source collimator.

Methods: The Monte Carlo model was developed using the Geant4 toolkit (release 10.6). The dual-focal-spot x-ray tube (Model MXR 225/22, COMET) was employed as the excitation source for XRF imaging. The computational x-ray source was filled with vacuum, and included a virtual monoenergetic electron gun (to simulate 2-cm tungsten filament), a rectangle W-target cut at 20 degree angle, and a 0.8-mm beryllium window (inherent filter). The electron beam hit the W-target with a spot size of 5.5 mm to mimic the operation conditions of the benchtop x-ray tube for XRF. The physics list was derived from PENELOPE, where the photoelectric effect, Compton Scattering, Rayleigh scattering, Coulomb scattering, ionization, multiple scattering, and bremsstrahlung were included. Small adjustments were made to the collimation and filtration thickness and material type until the produced energy spectrum was best suited for K-shell XRF. Once the best collimation and filtration were determined, the collimator was 3D printed as a steel shell and filled with lead, and the filter was constructed.

Results: For the x-ray tube without filtration and collimation, the simulated energy spectrum matched well with the reference spectrum, especially on the L- and K-edge characteristic energies of tungsten. Compared with Aluminum, Copper, and lead, Tin demonstrates more suitable for K-shell XRF imaging, with less background in the characteristic photon region (67-69 keV) for Gold.

Conclusion: Monte Carlo computational model of x-ray tubes can expedite the process of XRF system development and optimization, which saves costs, time and labor. This model will be expanded for future x-ray fluorescence computed tomography imaging.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email