Atlas Registration for Straight Trajectory Planning in Medical Images
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately plan a straight trajectory in medical images of anatomical structures without distorting the trajectory during image registration, which is crucial for procedures like spinal planning.
Innovation Solution
The method involves dividing an atlas containing a straight trajectory into restricted and unrestricted areas, applying a restricted transformation to maintain the trajectory's straightness and an unrestricted transformation to the rest of the atlas, ensuring the straightness of the trajectory is preserved during registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If an unrestricted transformation is applied to the entire atlas during image registration, then the atlas can be flexibly matched to the medical image, but the straight trajectory in the atlas becomes distorted and curved
Solution Approach 1:
The atlas is divided into two distinct regions: a first region containing the straight trajectory and a second region containing other anatomical structures. This segmentation allows different transformation rules to be applied to different parts of the atlas, enabling the trajectory to remain straight while other areas are freely transformed to match the medical image.
Solution Approach 2:
Different transformation qualities are applied to different regions of the atlas. The first region (containing the trajectory) undergoes a restricted transformation that preserves straightness, while the second region undergoes an unrestricted transformation that maximizes matching flexibility. This local differentiation resolves the contradiction between maintaining trajectory straightness and achieving flexible atlas matching.
2Shape
If a restricted transformation is applied to maintain trajectory straightness, then the trajectory remains straight, but the atlas cannot be optimally matched to the medical image
Solution Approach 1:
By segmenting the atlas into a first region (trajectory-containing) and a second region (other structures), the invention enables the first region to maintain straightness through restricted transformation while the second region achieves optimal matching accuracy through unrestricted transformation. The segmented approach allows both requirements to be satisfied simultaneously in their respective regions.
Solution Approach 2:
The invention applies different transformation qualities locally: restricted transformation in the first region to preserve trajectory straightness, and unrestricted transformation in the second region to maximize matching accuracy. This local quality differentiation resolves the contradiction by allowing each region to have the transformation properties it needs.
3Device complexity
If the entire atlas is transformed uniformly, then the transformation process is simple, but the trajectory becomes curved and anatomical structures are distorted
Solution Approach 1:
The atlas is segmented into a first region and a second region, allowing the transformation process to treat them differently. This segmentation increases transformation reliability by preserving trajectory integrity in the first region while allowing flexible matching in the second region, outweighing the increased complexity of managing two transformation types.
Solution Approach 2:
Different transformation qualities are applied locally to different regions: restricted transformation for the trajectory region to ensure integrity, and unrestricted transformation for other regions to enable proper matching. This local quality approach resolves the contradiction between process simplicity and trajectory reliability.
Data Source
AI summary
Disclosed is a computer-implemented method of planning a straight trajectory in a medical image of an anatomical structure. A template of the straight trajectory is given in an atlas and the atlas is registered with the medical image. The atlas is divided into a first area and one or more second areas, wherein the first area comprises the straight trajectory and the one or more second areas do not comprise any part of the straight trajectory. When the transformation of the atlas onto the medical image is calculated, the first and second areas are treated differently. The transformation of the first area maintains the straight shape of the trajectory and is thus restricted. The transformation of the one or more second areas is unrestricted and thus allows any kind of image registration. The template of the transformed atlas is the applied to the medical image.