Machine Learning Seminar[In-Person]: Dr. Vijay Sharma
University of Maryland School of Medicine
Location
Mathematics/Psychology : 412
Machine Learning Seminar[In-Person]: Dr. Vijay Sharma – Online Event
Date & Time
April 12, 2024, 1:00 pm – 2:00 pm
Description
Title: Status of Prompt Gamma MC simulations for proton therapy.
ABSTRACT: Prompt gamma (PG) imaging has emerged as a promising technique for real-time verification of proton beam range, thereby enhancing the precision of proton therapy. This study focuses on the development of a Monte Carlo simulation package aimed at investigating the energy spectra and yields of prompt gamma emissions to facilitate the optimization and implementation of PG imaging.
Utilizing the Geant4 Monte Carlo toolkit, we designed a simulation package to analyze the characteristic emission of PG lines during proton therapy delivery. The parameters defining the proton pencil beam profile, including shape, lateral spread, size, energy, and energy spread, were configured based on the treatment planning system. Patient anatomy was modeled using CT images with a resolution of approximately 0.97 mm x 0.97 mm x 3 mm, with material assignments in each voxel derived from Hounsfield Unit (HU) values. To ensure dosimetric accuracy and prompt gamma data sensitivity, the interaction step size was reduced to 0.1 mm at the microscopic level.
The simulation package successfully examines PG emission lines and their sources, identifying a favorable energy window between 3 MeV to 8 MeV. Analysis indicates that proton-induced reactions predominantly occur along the path of the pencil proton beams, while neutron-induced reactions significantly contribute from deeper regions within the patient's anatomy. Furthermore, the generation of PG emission lines gated spectra across various materials provides valuable insights into understanding PG production yield within the human body.
The simulation results revealed the potential of PG emission for real-time, in-vivo verification of proton therapy. The developed Monte Carlo tool serves as a robust platform for further investigation and optimization of PG imaging techniques, promising advancements in the field of proton therapy delivery and treatment verification.
Tags: