Perspective Image Alignment for Radiotherapy Target Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
It is challenging to accurately position a target for radiotherapy treatment due to significant posture differences between planning and treatment times, making it difficult to designate corresponding points on images captured at different times.
Innovation Solution
An image processor is used to generate and update perspective images based on three-dimensional volume data, acquire corresponding points on these images, and calculate differences in position to align the target accurately, incorporating a system that adjusts image directions and calculates rotation and translation matrices to refine positioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If corresponding points are designated on images captured at planning time and treatment time, then target positioning can be performed, but it becomes difficult when there is a large difference in posture between the two times
Solution Approach 1:
The patent generates a synthetic image by projecting three-dimensional information onto a two-dimensional plane, creating a virtual copy of the target's appearance at the planning time. This synthetic image is then overlaid with the actual image captured at treatment time, allowing corresponding points to be designated on the synthesized representation rather than requiring direct matching between vastly different postures.
Solution Approach 2:
The patent introduces three-dimensional volume data as an intermediary between the planning image and treatment image. By representing the target in three-dimensional space and projecting it to create a synthetic two-dimensional image, the system establishes a common reference frame that mediates the comparison between images taken at different times with different postures.
2Manufacturing precision
If three-dimensional volume data is used to generate perspective images, then alignment precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex mechanical positioning and alignment mechanisms with computational methods. Instead of using physical adjustment mechanisms to align images, the system uses three-dimensional volume data and projection algorithms to synthetically generate and align images, substituting mechanical complexity with computational processing.
Solution Approach 2:
The patent transitions from two-dimensional image processing to three-dimensional volume data processing. By representing the target in three-dimensional space and performing projections, the system achieves more precise alignment by utilizing the additional spatial dimension, allowing for accurate positioning even when two-dimensional images show large posture differences.
Data Source
AI summary
According to some embodiments, an image processor includes a perspective image generator, a perspective image acquirer, a corresponding point acquirer, a first calculator and a display image generator. The first calculator updates the position of the one or more first corresponding point on the first perspective image, based on a difference in position between the first and second perspective images. The display image generator generates a first display image including the first and second perspective images, the one or more first corresponding points updated on the first perspective image, and the one or more second corresponding points on the second perspective image. The perspective image generator changes the first direction into a third direction based on the difference in position between the first and second perspective images, and generates an updated one, viewed in the third direction, of the first perspective image.