On-Demand Dose-Volume Kernel Generation via Segmented Radial Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining the expected absorbed dose distribution in radiation therapy planning are time-intensive and lack flexibility, requiring pre-calculated dose-volume kernels for specific voxel sizes and isotopes, which is not suitable for on-demand generation during the planning process.
Innovation Solution
A radiation planning system that generates dose-volume kernels on-demand using a template function and radial dose distribution, allowing for quick calculation of expected absorbed dose distributions based on user-input voxel size and isotope type, reducing computational effort and storage needs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If pre-calculated dose-volume kernels are used for specific voxel sizes and isotopes, then calculation accuracy is improved, but computational time and storage requirements increase significantly
Solution Approach 1:
The dose-volume kernel is segmented into two independent components: a radial dose distribution component (isotope-specific) and a geometric template function component (voxel-size specific). This segmentation allows each component to be pre-calculated and stored separately, then combined on-demand through simple multiplication, avoiding the need to pre-calculate complete kernels for all voxel sizes and isotopes combinations.
Solution Approach 2:
The invention changes the parameter representation from complete 3D kernel arrays to 1D radial dose distribution functions and 1D geometric template functions. This parameter transformation reduces the data volume dramatically while maintaining the ability to generate accurate dose-volume kernels for any voxel size and isotope combination by combining the transformed parameters.
2Measurement precision
If pre-calculated dose-volume kernels are stored for different voxel sizes and isotopes, then calculation accuracy is improved, but storage space requirements increase
Solution Approach 1:
The storage requirement is segmented by separating universal components from specific components. The radial dose distribution depends only on the isotope type, while the geometric template function depends only on the voxel size. This segmentation allows universal radial dose distributions to be stored once per isotope, eliminating redundant storage across different voxel sizes.
Solution Approach 2:
The radial dose distribution function serves as a universal component that can be combined with multiple different geometric template functions to create dose-volume kernels for various voxel sizes and isotope combinations. This multi-functionality allows a single pre-calculated radial dose distribution to serve multiple purposes, dramatically reducing total storage requirements.
3Measurement precision
If Monte-Carlo techniques are used to pre-calculate dose-volume kernels, then calculation accuracy is improved, but the process becomes time-intensive and not suitable for on-demand generation
Solution Approach 1:
The radial dose distribution and geometric template functions are pre-calculated using accurate Monte-Carlo techniques during system setup. These pre-calculated components are then stored and can be combined instantly on-demand during treatment planning, achieving both high accuracy and fast on-demand generation without repeating the time-intensive Monte-Carlo calculations.
Solution Approach 2:
Instead of generating complete dose-volume kernels from scratch using Monte-Carlo simulations for each query, the system creates copies of pre-calculated radial dose distribution functions and geometric template functions, then combines them through simple mathematical operations. This copying approach maintains accuracy while dramatically reducing computation time for on-demand generation.
Data Source
AI summary
A radiation planning system includes a dose volume kernel determiner (122) and an expected absorbed dose determiner (124). The dose volume kernel determiner (122) generates a dose volume kernel for each of a plurality of voxels in a dose calculation grid. Each of the dose volume kernels is based on a radial dose distribution and a template function for a particular voxel size. The expected absorbed dose determiner (124) determines an expected absorbed dose distribution for each of the plurality of voxels based on the dose volume kernel.