Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Kanru_dissertation_final.pdf (17.84 MB)

ETD Abstract Container

Abstract Header

Monte Carlo Modeling of Virtual Multi-Featured Single Photon Source and High-Definition Multileaf Collimator for Modern Medical Linear Accelerators

Xie, Kanru
http://rave.ohiolink.edu/etdc/view?acc_num=toledo1639920165266538

Abstract Details

2021, Doctor of Philosophy, University of Toledo, Physics.
Linear accelerator (Linac) modeling is a common application of Monte Carlo (MC) method in medical physics. While the general-purpose Monte Carlo N-particle Code (MCNP) is considered one of the most powerful and flexible MC codes for radiation particle transports and interactions, its strength and flexibilities have not yet been fully explored in the area of Linac modeling. The main purpose of this dissertation work is to develop and validate a Linac head modeled in MCNP5 in the absence of detailed field independent Linac head component information and without phase space file (phsp) source cards. The significance of having this digitized MC Linac model is in providing efficiency and flexibility of doing research on our clinical implemented Linac machines. Linac modeling work consists of two separate parts: (1) model of the geometries of Linac head components; (2) model of the irradiation beam. The powerful geometry package of MCNP was utilized to build the geometries of the Linac head components of our Varian Edge Linac machine. Construction of the high definition multileaf collimator (HD-MLC) was the core work in this part. A total of 120 pieces of leaves of six different types in two opposite position leaf banks were modeled. Each leaf was constructed as a single cell in MCNP5 with 25 surfaces described to precisely define all the intricate features of a leaf including tongue-and-groove, supporting rails, triple-rounded tip and screw holes. The coordinates of the two big side surfaces of a leaf were carefully calculated to form a tiny angle in order to make the leaves have a physical divergence causing the rays to trace back to the target. Material of all collimators were set to be tungsten alloy with density of 18 g/cm3. Photon flux at 100 cm source-to-surface distance were tallied to verify the positions of collimators including jaws and MLCs. For the irradiation beam, single virtual photon source models of 6FFF and 6MV photon beams featuring angular distribution, angular dependent energy spectrum, and oval shape 2D spatial distribution were developed. Angular distribution and spectrum of the particles were extracted from a published phase space file (phsp) and modified in the modeling process to create a photon source with an angular dependent spectrum and the adjusted spatial distribution. To validate the model, energy deposition tallies for percent depth dose (PDD) and profiles of multiple field sizes and depths in water phantom were simulated and compared to the measured data (most importantly, PDD of 1010 cm2 field size, both inline and crossline profiles of 1010 and 3022 cm2 field size at 10 cm depth). In the absence of proprietary data from manufacturer, trial-and-error process was applied to tweak the virtual source parameters repeatedly many times until the simulation result showed good agreement with the measured data. An in-house-developed software package with graphical user interface (GUI) was created to assist the modeling work. Simulations of multiple field sizes defined by jaws or MLCs showed excellent accuracy of collimator positions. In terms of dosimetry, 99% of the points in simulated PDDs of both 6FFF and 6MV beams have dose differences less than 1% compared to measured data except the surface doses at 1-2 mm depth. Profiles of 6FFF beam have 100% gamma passing rate (2%/2mm criteria), for both inline and crossline profiles. Profile modeling of 6MV did not demonstrate the same high quality as the 6FFF model, where gamma test failed only in the out-of-field regions of the largest field size. Output factors of the conventional treatment field sizes in both beams showed good agreements with the machine commissioning data. The model was then used for some application projects including dosimetry verification, small fields output factors, development of a novel solid phantom, etc.. Preliminary results of these applications will be presented in this work. We conclude that a new method of a concise and flexible MC Linac model was validated in MCNP5, showing accuracy and reliability for medical physics research and dosimetry verification uses.
E. Ishmael Parsai (Advisor)
Diana Shvydka (Committee Member)
Nicholas Sperling (Committee Member)
Aniruddha Ray (Committee Member)
Yanfa Yan (Committee Member)
182 p.
Physics; Radiation
medical physics, Monte Carlo, Linac, beam modeling, multileaf collimator

Recommended Citations

Citations

  • Xie, K. (2021). Monte Carlo Modeling of Virtual Multi-Featured Single Photon Source and High-Definition Multileaf Collimator for Modern Medical Linear Accelerators [Doctoral dissertation, University of Toledo]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1639920165266538

    APA Style (7th edition)

  • Xie, Kanru. Monte Carlo Modeling of Virtual Multi-Featured Single Photon Source and High-Definition Multileaf Collimator for Modern Medical Linear Accelerators . 2021. University of Toledo, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=toledo1639920165266538.

    MLA Style (8th edition)

  • Xie, Kanru. "Monte Carlo Modeling of Virtual Multi-Featured Single Photon Source and High-Definition Multileaf Collimator for Modern Medical Linear Accelerators ." Doctoral dissertation, University of Toledo, 2021. http://rave.ohiolink.edu/etdc/view?acc_num=toledo1639920165266538

    Chicago Manual of Style (17th edition)

toledo1639920165266538
104
© 2021, all rights reserved.
This open access ETD is published by University of Toledo and OhioLINK.