Dual Model Dose Calculation for Radiation Treatment Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radiation treatment planning methods face challenges in efficiently and quickly generating high-quality treatment plans due to the complexity of beam geometries and intensities, requiring significant computational resources and time, especially in urgent cases like cancer treatment.
Innovation Solution
The methodology employs two dose prediction models: a faster, less accurate model for initial calculations and a more accurate, resource-intensive model for refined plans, allowing for quick generation of additional treatment plans by projecting new fluence maps onto existing ones and selecting the appropriate model based on residual values to minimize computational demands.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a more accurate dose prediction model is used, then manufacturing precision is improved, but productivity deteriorates
Solution Approach 1:
The patent segments the dose prediction process into two distinct models: a first dose prediction model for rapid initial calculations and a second dose prediction model for refined accurate calculations. This segmentation allows the system to use the appropriate level of computational complexity for each stage of plan generation, improving overall efficiency while maintaining accuracy where needed.
Solution Approach 2:
The system dynamically changes parameters (specifically computational resources and model selection) based on the treatment planning stage and requirements. By adjusting which model is used at different stages, the system optimizes the balance between speed and accuracy without being constrained to a single fixed approach.
2Manufacturing precision
If computational resources are increased, then manufacturing precision is improved, but loss of energy increases
Solution Approach 1:
The patent divides the computational workload into two segments: initial dose calculations using a resource-efficient first model, and refined dose calculations using a more resource-intensive second model only when necessary. This segmentation reduces overall energy consumption while maintaining the required level of accuracy.
Solution Approach 2:
The system applies partial action by using the simpler first dose prediction model for the majority of calculations where full accuracy is not critical, and reserves the more computationally intensive second model for specific cases requiring higher precision. This avoids excessive energy consumption while maintaining sufficient accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A first dose calculated using a first set of fluence maps and a first (faster) dose prediction model is accessed 602. A second fluence map is accessed 604. The second fluence map is projected onto the first set of fluence maps to determine a set of scalars and a residual value 606. When the residual value satisfies a criterion, a second dose is calculated using the first dose prediction model, the set of scalars, and the second fluence map 608. When the residual value does not satisfy the criterion, the second dose is calculated using a second (more accurate) dose prediction model and the second fluence map 608.