Representing Range Compensators with Computational Geometry in TOPAS

  • Slides: 13
Download presentation
Representing Range Compensators with Computational Geometry in TOPAS Forrest Iandola 1, 2 and Joseph

Representing Range Compensators with Computational Geometry in TOPAS Forrest Iandola 1, 2 and Joseph Perl 1 SLAC National Accelerator Laboratory 2 University of Illinois at Urbana-Champaign 1

Overview • • • Medical Physics in 30 Seconds Introduction to TOPAS What is

Overview • • • Medical Physics in 30 Seconds Introduction to TOPAS What is a range compensator? Subtraction Solid geometry Modeling compensators with Subtraction Solids Approximation with polyhedrons for (potentially) faster performance 2 Forrest Iandola Computational Geometry Compensators

Medical Physics in 30 Seconds • Goal: kill cancer with radiation • Deliver radiation

Medical Physics in 30 Seconds • Goal: kill cancer with radiation • Deliver radiation with protons, photons, other particles, or ions • Monte Carlo simulation of proton therapy beams is an up-and-coming field 3 Forrest Iandola Computational Geometry Compensators

Introduction to TOPAS (Tool for Particle Simulation) • TOPAS = Monte Carlo simulation of

Introduction to TOPAS (Tool for Particle Simulation) • TOPAS = Monte Carlo simulation of radiation therapy beamlines • User can easily customize beamline for specific treatment facilities • Uses Geant 4 for the “real” physics 4 Forrest Iandola Computational Geometry Compensators

What is a Range Compensator? • A radiation therapy beamline collimates the beam and

What is a Range Compensator? • A radiation therapy beamline collimates the beam and produces a specific energy spread • Range compensator produces an energy spread • Construction: drill a number of holes out of a cylinder of lucite • Each drill hole may have a unique depth • “The thickness of the Lucite [plastic] will proportionally reduce the depth [energy] of the protons” 1 1 http: //neurosurgery. mgh. harvard. edu/protonbeam/ 5 Forrest Iandola Computational Geometry Compensators

What is a Range Compensator? 6 Forrest Iandola Computational Geometry Compensators

What is a Range Compensator? 6 Forrest Iandola Computational Geometry Compensators

Subtraction Solids • Geant 4 supports boolean solid combinatorial geometry – Subtraction solids –

Subtraction Solids • Geant 4 supports boolean solid combinatorial geometry – Subtraction solids – Union solids • It’s as simple as new. Solid = Solid 1 • Overlap among subtracted solids is acceptable • Solids can be recursively subtracted 7 Solid 2 Forrest Iandola Computational Geometry Compensators

Compensator with Subtraction Solids Compensator comprised of a big. Cylinder with n holes subtracted:

Compensator with Subtraction Solids Compensator comprised of a big. Cylinder with n holes subtracted: new. Solid 1 = big. Cylinder - small. Cylinder 1 new. Solid 2 = new. Solid 1 - small. Cylinder 2 … Compensator = new. Solid(n-1) - small. Cylinder(n) 8 Forrest Iandola Computational Geometry Compensators

Compensator with Subtraction Solids 9 Forrest Iandola Computational Geometry Compensators

Compensator with Subtraction Solids 9 Forrest Iandola Computational Geometry Compensators

Approximation for Performance Gains • Approximate the drill holes with a collection of hexagons

Approximation for Performance Gains • Approximate the drill holes with a collection of hexagons • Lack of overlap among hexagons allows us to model all hexagons as a single polyhedron • Future work: evaluate performance benefits (and accuracy reduction) with polyhedron method 10 Forrest Iandola Computational Geometry Compensators

Approximation for Performance Gains Subtraction Solid Polyhedron 11 Forrest Iandola Computational Geometry Compensators

Approximation for Performance Gains Subtraction Solid Polyhedron 11 Forrest Iandola Computational Geometry Compensators

Approximation for Performance Gains • Future work: evaluate performance benefits (and accuracy reduction) with

Approximation for Performance Gains • Future work: evaluate performance benefits (and accuracy reduction) with polyhedron method 12 Forrest Iandola Computational Geometry Compensators

Acknowledgements • Harald Paganetti (Massachusetts General Hospital and Harvard University) • Jan Schuemann (Massachusetts

Acknowledgements • Harald Paganetti (Massachusetts General Hospital and Harvard University) • Jan Schuemann (Massachusetts General Hospital and Harvard University) • Jungwook Shin (UC San Francisco) • Bruce Faddegon (UC San Francisco) • DOE and NIH for generous support Contact: forrest@slac. stanford. edu 13 Forrest Iandola Computational Geometry Compensators