Speaker
Dr
Heather Chen-Mayer
(NIST)
Description
Prompt gamma activation analysis (PGAA) is an elemental analysis technique using gamma rays emitted upon neutron capture by nuclei contained within a sample. It provides quantitative information for the elemental composition while maintaining the sample integrity. The primary limitation of PGAA, however, is that it is a bulk technique giving average concentrations, without spatial information. This work explores Compton Imaging tomography of the prompt gamma rays, with the goal of developing a new multi-dimensional method that can provide both quantitative and spatial information for the elemental distributions within a sample.
We report preliminary work on computer simulation of Compton imaging of prompt gamma emissions and reconstruction to localize the source origin. Simulation was conducted using the Monte Carlo code GEANT4 on a series of simple systems to evaluate the feasibility of applying Compton imaging tomography to the element of interest at a scale of interest for elemental analysis. The reconstructions were performed using back projection and a statistical method. A stack of alternating disks of Ti and H (in the form of water) of 2 cm in diameter and 0.5 mm thick was placed perpendicularly to a set of two detector planes made with CZT material. Prompt gamma rays produced by a beam of incident neutrons were tracked, with those scattered from the first detector plane and absorbed by the second recorded.
The GEANT4 simulated energy histograms were windowed by the main Ti prompt gamma peak at 1381 keV and H peak at 2223 keV to produce energy and position information as input to the reconstruction code to recover spatial information at the source origin. The Ti and H emission images were recovered and can be shown separately or as a composite. The result from the back-projection method and the statistical reconstruction method are compared. The 5.5-mm spacing of the disks is fully resolvable. In addition, it is worthwhile noting that the statistical reconstruction was about 50 times faster than the back-projection method for this example. The statistical reconstruction was done in full 3D, whereas the back-projection was performed slice by slice.
Further refinement of the model to reflect experimental setup will be carried out. The result of this study will facilitate improvement of the reconstruction process for the data obtained earlier on the NCNR PGAA instrument.
Primary author
Mr
Benjamin Riley
(University of Kentucky)
Co-authors
Dr
Danyal Turkoglu
(NIST)
Dr
Haijian Chen
(University of Maryland School of Medicine)
Dr
Heather Chen-Mayer
(NIST)
Prof.
Jerimy Polf
(University of Maryland School of Medicine)
