Image Processing Device with Iterative Vertebral Landmark Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for landmark detection in images, particularly for vertebral bodies, struggle to accurately derive corner points using offset regression due to ambiguous features and distances from reference points.
Innovation Solution
An image processing method involving a processor that performs first and second processing steps to iteratively derive offsets from reference points to candidate points until a predetermined condition is satisfied, using a trained derivation model to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If offset regression is used to derive corner points from the center of vertebral body, then the detection process is simplified, but the accuracy of landmark detection deteriorates due to ambiguous features and distance from reference points
Solution Approach 1:
The detection process is segmented into multiple stages: first detecting the center of vertebral body as a reference point, then using offset regression to derive corner points. When accuracy is insufficient, the system performs iterative refinement by detecting new reference points near candidate corner points and re-calculating offsets, thus dividing the detection task into manageable segments that progressively improve accuracy
Solution Approach 2:
The system performs preliminary detection of the vertebral body center and initial corner points using offset regression. Based on the accuracy of this preliminary detection, the system decides whether to perform additional iterative refinement steps, thus preparing the detection framework in advance while maintaining flexibility for accuracy improvement
2Measurement precision
If iterative refinement is performed to improve landmark detection accuracy, then detection precision improves, but processing time increases
Solution Approach 1:
The system performs iterative refinement only when necessary - specifically when the initial offset regression detection does not achieve sufficient accuracy. The iteration continues for a limited number of times (N times) or until a predetermined condition is satisfied, thus applying partial action rather than always performing full iterative refinement, balancing accuracy improvement with processing time consumption
Data Source
AI summary
A processor performs first processing of acquiring an offset from a reference point of a structure included in an image to a candidate point of a target point related to the reference point, and repeatedly performs second processing of acquiring a new offset to a new candidate point of the target point in which the candidate point of the target point derived based on the offset is used as a new reference point, N times (N≥1) until a predetermined condition is satisfied, to derive the target point.