Medical Image Processing Apparatus for Protruding Shape Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical image processing systems face challenges in accurately detecting protruding shapes within three-dimensional models of living tissues in body cavities, particularly in differentiating between visible and invisible regions and identifying lesions like polyps.
Innovation Solution
The system estimates a three-dimensional model from a two-dimensional image, calculates shape feature values, and extracts voxel groups based on predetermined shapes, including convex and roof shapes, to detect protruding shapes by analyzing visible and invisible regions and their spatial relationships.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If three-dimensional model estimation is performed from two-dimensional images to detect protruding shapes, then detection capability is improved, but differentiation between visible and invisible regions becomes difficult
Solution Approach 1:
The patent applies dimensionality change by transitioning from two-dimensional image data to three-dimensional model representation. The three-dimensional model estimating section reconstructs the internal structure of living tissue by generating voxel-based 3D models from multiple 2D images, enabling detection of protruding shapes in the depth dimension that cannot be observed in planar images alone.
Solution Approach 2:
The patent segments the three-dimensional model into distinct voxel groups representing different tissue regions. The visible region detecting section identifies voxels corresponding to observable tissue surfaces, while the invisible region detecting section identifies voxels representing internal or occluded structures. This segmentation allows clear differentiation between visible and invisible regions within the 3D model.
2Measurement precision
If shape feature values are calculated for all voxels to improve detection accuracy, then measurement precision is improved, but processing complexity increases
Solution Approach 1:
The patent applies local quality by calculating shape feature values selectively for specific voxel groups rather than uniformly for all voxels. The shape feature value calculating section computes curvature and other geometric features only for voxels identified as part of potential protruding shapes or lesion areas, reducing computational complexity while maintaining detection accuracy in critical regions.
Solution Approach 2:
The patent implements partial action by focusing computational resources on calculating shape feature values for a subset of voxels that are most likely to contain lesions. Instead of processing the entire three-dimensional model, the system concentrates analysis on regions with abnormal geometric characteristics, achieving high detection accuracy with reduced processing complexity.
3Reliability
If multiple shape types are extracted to improve lesion identification, then detection reliability is improved, but device complexity increases
Solution Approach 1:
The patent applies universality by designing a multi-functional three-dimensional shape extracting section that can identify multiple types of protruding shapes (convex, concave, cylindrical, spherical) using a unified computational framework. This single module performs various shape classification functions, improving lesion identification reliability across different polyp types without requiring separate dedicated systems for each shape category.
Data Source
AI summary
A medical image processing apparatus of the present invention has a three-dimensional model estimating section for estimating, based on an inputted two-dimensional image of an image of a living tissue within a body cavity, a three-dimensional model of the living tissue, a shape feature value calculating section for calculating shape feature values of respective voxels included in the three-dimensional model of the living tissue, a three-dimensional shape extracting section for extracting a first voxel group whose three-dimensional model has been estimated as a predetermined shape, in the respective voxels included in the three-dimensional model of the living tissue, based on the shape feature values, and a protruding shape detecting section for detecting the first voxel group as a voxel group configuring a protruding shape in the three-dimensional model of the living tissue.