AI Radiation Dose Control for Patient Motion Adaptation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing radiation therapy methods struggle to effectively deliver precise doses to target tissues while minimizing exposure to surrounding healthy tissues, particularly due to patient movements and anatomical changes during treatment fractions.

Innovation Solution

A system utilizing an artificial intelligence agent trained through reinforcement learning to control radiation delivery parameters, accounting for patient motions and anatomical geometry changes, to optimize the delivery of radiation therapy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If radiation beams are delivered from multiple directions with shaped cross-sections to conform to target volume projections, then the precision of dose delivery to target tissue is improved, but the complexity of the radiation delivery system increases

Engineering Contradiction:
Improvedose delivery precisionVSAvoidradiation delivery system complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The system employs dynamic adaptation by using an AI agent that continuously monitors patient anatomy and adjusts radiation delivery parameters in real-time during treatment fractions. This allows the system to maintain high dose delivery precision while managing complexity through adaptive control rather than static pre-planning alone.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system implements feedback mechanisms where the AI agent receives continuous input about patient position and anatomy changes during treatment, then adjusts delivery parameters accordingly. This closed-loop control enables precise dose delivery while the AI autonomously manages the complexity of real-time adjustments.

Inventive Principle:
Principle #23Feedback

2Productivity

If the radiation delivery apparatus is operated automatically based on pre-planned control points and parameters, then the productivity of treatment delivery is improved, but the adaptability to patient motions and anatomical changes deteriorates

Engineering Contradiction:
Improvetreatment delivery efficiencyVSAvoidadaptability to patient motions
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The system transitions from static pre-planned delivery to dynamic adaptive delivery by employing an AI agent that continuously adjusts radiation parameters during treatment fractions based on real-time patient monitoring data, maintaining both efficiency and adaptability.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The AI agent autonomously monitors patient anatomy and adjusts delivery parameters without requiring manual intervention during treatment fractions. This self-service capability allows the system to maintain high productivity while adapting to patient motions and anatomical changes in real-time.

Inventive Principle:
Principle #25Self-service

3Object-affected harmful factors

If control points and radiation delivery parameters are optimized to minimize radiation exposure to critical structures, then the harmful effects on healthy tissue are reduced, but the device complexity and treatment planning time increase

Engineering Contradiction:
Improveradiation exposure to healthy tissueVSAvoidtreatment planning complexity
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The AI agent autonomously optimizes radiation delivery parameters during treatment fractions based on real-time patient monitoring, automatically adjusting beam parameters to minimize exposure to healthy tissues without requiring complex manual re-planning for each fraction.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system uses feedback from real-time patient monitoring to continuously adjust delivery parameters, enabling the AI to dynamically optimize dose distribution and minimize exposure to critical structures during treatment without increasing overall system complexity.

Inventive Principle:
Principle #23Feedback

4Ease of manufacture

If treatment plans are created based on stationary planning images, then the ease of treatment planning is improved, but the manufacturing precision of dose delivery deteriorates due to patient motions

Engineering Contradiction:
Improvetreatment planning easeVSAvoiddose delivery precision
Core Design Contradiction:
Ease of manufactureVSManufacturing precision

Solution Approach 1:

The system maintains simple stationary planning but adds dynamic adaptation during delivery through the AI agent, which continuously adjusts parameters based on real-time patient monitoring to compensate for motions, preserving both planning ease and delivery precision.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system performs preliminary treatment planning based on stationary images for ease of use, then employs the AI agent to make real-time adjustments during delivery to account for patient motions, combining the benefits of simple planning with precise adaptive delivery.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4126212B1A system for delivering radiation dose to a patient using an artificial intelligence agent
Publication Date: 2026.02.25 SIEMENS HEALTHINEERS INTERNATIONAL AG
  • EP4126212B1 patent drawingFigure 1~1A
  • EP4126212B1 patent drawingFigure 2
  • EP4126212B1 patent drawingFigure 3A~3B

AI summary

Methods and systems are provided which relate to the planning and delivery of radiation treatments by modalities which involve moving a radiation source along a trajectory relative to a subject while delivering radiation to the subject. An artificial intelligence (AI) agent trained using reinforcement learning (and/or some other suitable form of machine learning) is used to control the radiation delivery parameters in effort to achieve desired delivery of radiation therapy. In some embodiments, the AI agent selects suitable control steps (e.g. radiation delivery parameters for particular time steps), while accounting for patient motions, difference(s) in patient anatomical geometry and/or the like.