Image Processing Device Mapping Accuracy Overlapped Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for extracting and matching candidate points from image data to shape models often result in inappropriate selections, leading to partially or fully overlapped paths, which affects the accuracy of structure detection in medical imaging.
Innovation Solution
An image processing device and method that extracts candidate points from image data and selects corresponding points based on specific constraints to ensure each model label is mapped to only one candidate point, and each candidate point is mapped to only one model label, with paths determined to minimize cost, using a Dual Decomposition method to optimize the mapping and energy function.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional mapping methods are used to select corresponding points from candidate points to model labels, then the mapping process is simple, but the mapping accuracy deteriorates due to inappropriate selections and overlapped paths
Solution Approach 1:
The mapping process is divided into two independent phases: first determining which candidate points are mapped to model labels (mapping relationship), and second determining paths between mapped points. This segmentation allows the mapping accuracy to be improved by independently optimizing the mapping relationship selection, while the path determination handles the geometric constraints. The mapping relationship determination uses constraint (c) to prevent overlapped paths, resolving the contradiction between accuracy and complexity.
Solution Approach 2:
The mapping relationship between candidate points and model labels is determined in advance, before path determination. This preliminary action establishes a foundation for accurate mapping by selecting optimal candidate points based on evaluation functions and constraints, including the forward constraint that prevents future path overlaps. This preliminary determination of mapping relationships improves overall mapping accuracy while organizing the complexity into manageable stages.
2Reliability
If the set of candidate points is selected from feasible solutions with basic constraints only, then the selection process is fast, but the mapping appropriateness deteriorates due to inappropriate point selections at branch points
Solution Approach 1:
The forward constraint (constraint c) is applied in advance during mapping relationship determination to prevent inappropriate selections that would cause path overlaps. By evaluating potential mapping assignments against the forward constraint before finalizing the mapping, the system prevents erroneous mappings at branch points and other critical locations. This preliminary anti-action improves mapping appropriateness by eliminating problematic assignments before they can cause path overlap issues, while the efficient evaluation function maintains computational speed.
Solution Approach 2:
The forward constraint creates a feedback mechanism where the determination of mapping relationships considers future path determination requirements. The evaluation function incorporates information about potential path overlaps, providing feedback that guides the selection of appropriate candidate points. This feedback loop ensures that mapping decisions account for their impact on subsequent path determination, improving overall mapping appropriateness without requiring exhaustive computation.
3Measurement precision
If additional constraints are added to ensure non-overlapped paths, then the mapping accuracy improves, but the computational complexity increases
Solution Approach 1:
The constraint processing is segmented into two distinct parts: the forward constraint (constraint c) applied during mapping relationship determination, and the backward constraint applied during path determination. This segmentation allows each constraint type to be processed efficiently in its appropriate phase, improving detection precision through comprehensive constraint satisfaction while managing computational complexity by avoiding redundant constraint checking in each phase.
Solution Approach 2:
The forward constraint is applied preliminarily during mapping relationship determination, establishing a foundation that prevents path overlaps before path determination begins. This preliminary application of constraints reduces the computational burden during the subsequent path determination phase, as fewer invalid path configurations need to be evaluated. The preliminary constraint satisfaction improves detection precision while optimizing the overall computational workflow.
Data Source
AI summary
An image processing device includes a candidate point extracting unit configured to extract a plurality of candidate points belonging to a predetermined structure from image data, a shape model storing unit configured to store a shape model representing a known shape of the predetermined structure, the shape model being formed by a plurality of model labels having a predetermined connection relationship, and a corresponding point selecting unit configured to select a mapping relationship between the candidate points and the model labels from a set of candidate mapping relationships.