Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm

The CyberKnife Robotic RadioSurgery System is a robot controlled 6 MV linear accelerator based radiation delivery system with the linear accelerator attached to a six-axis robotic manipulator. Summation of all radiation beams creates a three-dimensional dose distribution within a patient. Each bea...

Full description

Bibliographic Details
Main Author: Clancey, Owen
Language:English
Published: 2011
Subjects:
Online Access:https://doi.org/10.7916/D854315R
id ndltd-columbia.edu-oai-academiccommons.columbia.edu-10.7916-D854315R
record_format oai_dc
spelling ndltd-columbia.edu-oai-academiccommons.columbia.edu-10.7916-D854315R2019-05-09T15:15:38ZDosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic AlgorithmClancey, Owen2011ThesesPhysicsComputer scienceMemeticsBiomedical engineeringRadiation dosimetryThe CyberKnife Robotic RadioSurgery System is a robot controlled 6 MV linear accelerator based radiation delivery system with the linear accelerator attached to a six-axis robotic manipulator. Summation of all radiation beams creates a three-dimensional dose distribution within a patient. Each beam's direction, weight, and collimator size affect its contribution to the dose distribution. Hence, the CyberKnife treatment planning problem is to select a set of beams that produce a desired dose distribution. With a dose-based objective function and user-supplied weighted, dose-volume goals, a memetic algorithm is used to solve the CyberKnife treatment planning problem. Before optimization begins, two thousand radiation beams are generated, and for each beam, dose-deposition coefficients are calculated for all optimization points within the target(s) and critical structures. Then, the memetic algorithm optimizes beam weights using global and local operators and problem-specific knowledge within an evolutionary computation framework. Concurrently, beams are pared down to emphasize promising regions of the solution space and to generate clinically deliverable treatment plans. Algorithmic analysis is two-fold: parameter analysis and comparison to MultiPlan, the only commercially available CyberKnife treatment planning software. Parameter analysis optimizes and justifies parametric choices given hardware, optimization time, and treatment time constraints analogous to clinical limitations. Thereafter, MultiPlan and the memetic algorithm generate ten treatment plans and are evaluated based upon dose-volume histograms, target dose homogeneity, target dose conformality, dosimetric success rates, total beam-on time or MU, and total number of beams. Analysis shows the memetic algorithm is equivalent or superior for all metrics, and given that MultiPlan is the only available CyberKnife treatment planning software, the memetic algorithm is a state-of-the-art CyberKnife dosimetric optimization method.Englishhttps://doi.org/10.7916/D854315R
collection NDLTD
language English
sources NDLTD
topic Physics
Computer science
Memetics
Biomedical engineering
Radiation dosimetry
spellingShingle Physics
Computer science
Memetics
Biomedical engineering
Radiation dosimetry
Clancey, Owen
Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
description The CyberKnife Robotic RadioSurgery System is a robot controlled 6 MV linear accelerator based radiation delivery system with the linear accelerator attached to a six-axis robotic manipulator. Summation of all radiation beams creates a three-dimensional dose distribution within a patient. Each beam's direction, weight, and collimator size affect its contribution to the dose distribution. Hence, the CyberKnife treatment planning problem is to select a set of beams that produce a desired dose distribution. With a dose-based objective function and user-supplied weighted, dose-volume goals, a memetic algorithm is used to solve the CyberKnife treatment planning problem. Before optimization begins, two thousand radiation beams are generated, and for each beam, dose-deposition coefficients are calculated for all optimization points within the target(s) and critical structures. Then, the memetic algorithm optimizes beam weights using global and local operators and problem-specific knowledge within an evolutionary computation framework. Concurrently, beams are pared down to emphasize promising regions of the solution space and to generate clinically deliverable treatment plans. Algorithmic analysis is two-fold: parameter analysis and comparison to MultiPlan, the only commercially available CyberKnife treatment planning software. Parameter analysis optimizes and justifies parametric choices given hardware, optimization time, and treatment time constraints analogous to clinical limitations. Thereafter, MultiPlan and the memetic algorithm generate ten treatment plans and are evaluated based upon dose-volume histograms, target dose homogeneity, target dose conformality, dosimetric success rates, total beam-on time or MU, and total number of beams. Analysis shows the memetic algorithm is equivalent or superior for all metrics, and given that MultiPlan is the only available CyberKnife treatment planning software, the memetic algorithm is a state-of-the-art CyberKnife dosimetric optimization method.
author Clancey, Owen
author_facet Clancey, Owen
author_sort Clancey, Owen
title Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
title_short Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
title_full Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
title_fullStr Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
title_full_unstemmed Dosimetric Optimization Method for CyberKnife Robotic RadioSurgery System Using a Memetic Algorithm
title_sort dosimetric optimization method for cyberknife robotic radiosurgery system using a memetic algorithm
publishDate 2011
url https://doi.org/10.7916/D854315R
work_keys_str_mv AT clanceyowen dosimetricoptimizationmethodforcyberkniferoboticradiosurgerysystemusingamemeticalgorithm
_version_ 1719046759814004736