Uncertainty-Guided Beam Control for Radiation Targeting

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing radiation therapy systems fail to account for uncertainty in target location due to limitations in image acquisition/reconstruction and subjective annotation, leading to inaccurate treatment plans that may deliver radiation to non-target tissue or miss target tissue.

Innovation Solution

A radiation therapy system that utilizes uncertainty values associated with voxels to generate beam control data, accounting for probabilities that voxels belong to the target, thereby improving treatment planning, organ tracking, and dose accumulation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual or automated annotation is used to define target boundaries, then treatment planning can be performed, but uncertainty in target location leads to inaccurate radiation delivery

Engineering Contradiction:

Improvetarget location accuracyVSAvoidtreatment plan accuracy

Core Design Contradiction:

Measurement precisionVSReliability

Solution Approach 1:

The patent transforms binary target/non-target classification into a probabilistic framework by introducing uncertainty parameters (0-1 values) for each voxel. This parameter change allows the system to quantify and propagate annotation uncertainty through the treatment planning process, improving both measurement precision and reliability by accounting for boundary ambiguities.

Inventive Principle:

Principle #35Parameter changes

2Ease of operation

If radiation beams are directed based on binary target outlines, then treatment delivery is simplified, but healthy tissue may receive unnecessary radiation exposure

Engineering Contradiction:

Improvebeam control simplicityVSAvoidradiation exposure to healthy tissue

Core Design Contradiction:

Ease of operationVSObject-affected harmful factors

Solution Approach 1:

The patent applies local quality by varying beam control parameters based on spatially-dependent uncertainty values. Voxels with high uncertainty (boundary regions) receive different beam control treatment compared to voxels with low uncertainty (core regions), allowing precise modulation of radiation delivery to minimize healthy tissue exposure while maintaining operational feasibility.

Inventive Principle:

Principle #3Local quality

3Measurement precision

If uncertainty values are calculated for each voxel, then treatment accuracy improves, but computational complexity increases

Engineering Contradiction:

Improvevoxel classification accuracyVSAvoidcomputational processing complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the target volume into discrete voxels and calculates uncertainty values independently for each voxel based on image intensity characteristics. This segmentation approach distributes the computational burden across many simple, parallel calculations rather than requiring complex global optimization, thereby improving measurement precision while managing computational complexity through divide-and-conquer.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS20260000915A1Apparatus and method of generating beam control data based on imaging uncertainty

Publication Date: 2026.01.01 ACCURAY LLC

AI summary

An apparatus includes a memory to store volumetric data representing a volumetric image of an anatomical region having a volume of interest (VOI), and a processing device operatively coupled to the memory. The processing device is to determine, based on the volumetric data, several voxels of the volumetric image. Each voxel has a respective uncertainty value representing a probability that the voxel belongs to the VOI. The processing device is further to generate beam control data to direct a treatment beam relative to the VOI based on the uncertainty values to meet a predetermined quality metric.