MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel
  1. Home
  2. Meeting Program
  3. All Sessions
  4. All Imaging ePoster Campus
  5. A GATE-Driven Monte Carlo Framework for Modeling Radiation Detection System

A GATE-Driven Monte Carlo Framework for Modeling Radiation Detection System

P Baturin1*, M Myronakis2 , Y Hu3 , D Shedlock4 , R Berbeco5 , J Star-Lack6 , (1) Varian Medical Systems, Palo Alto, California, (2) Department of Radiation Oncology, Brigham and Women's Hospital, Dana-Farber, Boston, MA, (3) Dana Farber Cancer Institute, Boston, MA, (4) Varian Medical Systems, Palo Alto, California, (5) Department of Radiation Oncology, Brigham and Women's Hospital, Dana-Farber, Boston, MA, (6) Varian Medical Systems, Palo Alto, CA

Presentations

(Monday, 7/30/2018) 4:30 PM - 5:30 PM

Room: Exhibit Hall | Forum 1

Purpose: Monte Carlo (MC) radiation and optical transport simulators are the de facto tool for modeling and characterization of radiation detection systems. Despite their flexibility and adaptability to user-specific tasks, the modeling of a compete radiation detection system is complex due to the large amount of parameterization required. In this work we report a simplified parameterization framework that makes the modeling of complex detectors practical.

Methods: Geant4 Application for Tomographic Emission (GATE) was chosen as the back-end platform for the framework which allows for modeling of an entire system including the x-ray source, imaging phantoms and a radiation detector along with relevant radiative and optical transport properties. We present an overview of the simulation architecture along with model examples of two flat panel-based high-efficiency portal imagers: 1) a 4-layer imager (MLI) with Gadolinium Oxy-Sulfide scintillators and 2) a single-layer imager (SLI) with a thick glass scintillator. Comparisons between simulated and measured modulation transfer functions (MTF), noise power spectra (NPS), and detective quantum efficiencies (DQE) were performed including computation of normalized root-mean square errors (NRMSE) and Pearson’s correlation coefficients.

Results: Execution time for the simulations ranged from 30 to 88 minutes on a single Xeon CPU. The MTF, NPS, and DQE correlation coefficients for the MLI were above 0.99 with the NRMSE being on the order of 10-2. The MTF and NPS Pearson’s coefficients of the SLI were above 0.99, while the DQE coefficient was 0.90. The NRMSE results were below 0.04 for MTF and NPS and was 0.19 for the DQE curves.

Conclusion: We have developed a user-friendly MC simulation toolkit that models entire radiation imaging systems. The toolkit was validated against experimental measurements. It is expected that the designed toolkit can be used to optimize and evaluate existent equipment and to facilitate the development of new imaging systems.

Funding Support, Disclosures, and Conflict of Interest: The work described was in part supported by a research grant from the National Cancer Institute (Award Number R01CA188446).

Keywords

Not Applicable / None Entered.

Taxonomy

Not Applicable / None Entered.

Contact Email

Additional Links