VMAT Collimator Trajectory Optimization via Segmented Geometric Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radiation therapy planning for Volumetric Modulated Arc Therapy (VMAT) faces challenges in achieving optimal fidelity to dose objectives while maintaining computational efficiency, often requiring a trade-off between parameter optimization and reduced complexity, leading to sub-optimal results due to the large number of adjustable parameters.
Innovation Solution
A two-step optimization process is introduced, where a geometric optimization is performed first to optimize collimator angles and gantry rotation speed without calculating radiation absorption profiles, followed by a main optimization using these initialized values to refine the radiation therapy plan, reducing computational complexity and improving fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a large number of parameters are optimized during VMAT planning, then the fidelity to dose objectives is improved, but the computational complexity and time required increase significantly
Solution Approach 1:
The patent segments the VMAT planning process into two distinct phases: a geometric optimization phase that optimizes collimator angles and gantry rotation speeds without radiation absorption calculations, and a main optimization phase that performs full dose calculation with fewer parameters. This segmentation allows complex geometric parameter optimization to be performed efficiently, followed by a simpler main optimization, thereby improving fidelity to dose objectives while managing computational complexity.
Solution Approach 2:
The patent applies preliminary action by performing geometric optimization of collimator angles and gantry rotation speeds before the main optimization phase. This preliminary optimization establishes good initial values for these parameters, which then serve as fixed or constrained parameters during the main optimization, reducing the search space and computational burden while improving the likelihood of achieving optimal dose distribution.
2Productivity
If collimator angle is set to a fixed value for all control points, then computational efficiency is improved, but the fidelity to dose objectives deteriorates
Solution Approach 1:
The patent applies dynamics by making the collimator angle a variable parameter that can be optimized during the geometric optimization phase, rather than fixing it to a constant value. The optimization determines optimal collimator angles for different control points based on geometric considerations, allowing the system to adapt collimator angles dynamically across the arc while maintaining computational efficiency through the two-phase approach.
3Measurement precision
If the number of control points is increased to accurately discretize the arc, then the precision of radiation delivery is improved, but the total number of parameters to optimize increases
Solution Approach 1:
The patent segments the optimization parameters into two groups: geometric parameters (collimator angles and gantry rotation speeds) that are optimized in the first phase, and radiation therapy parameters (MLC leaf positions and fluence) that are optimized in the main phase. This segmentation allows accurate discretization with many control points while managing parameter complexity through phased optimization.
Data Source
AI summary
In a continuous arc radiation therapy planning method for planning a radiation therapy session parameterized by a set parameters for control points (CPs) along at least one radiation source arc, a geometric optimization (40) is performed that does not include calculating radiation absorption profiles to generate optimized values for a sub-set of the parameters. After the geometric optimization, a main optimization (42) is performed that includes calculating radiation absorption profiles. The main optimization is performed with the sub-set of parameters initialized to the optimized values from the geometric optimization. The sub-set of parameters optimized by the geometric optimization may include collimator angle parameters for a multileaf collimator (MLC) (58). The geometric optimization may optimize a cost function comprising a sum over the CPs of a per-CP cost function dependent on a target-only region (62) defined as a planning target volume excluding any portion overlapping an organ at risk.