Bone Group Segmentation for Medical Image Registration Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image registration techniques in the medical field face challenges in accurately registering images with large bone motion differences or discontinuous deformations, leading to misregistration and instability, especially in cases where bones are cut or have undergone surgery.
Innovation Solution
An image processing apparatus that identifies and classifies bones into groups based on their associated body parts, performing registration for each group independently to generate a difference image, thereby stabilizing the registration process and avoiding discontinuous deformation issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If registration is performed between two three-dimensional images with large bone motion differences or discontinuous deformation, then the registration process becomes complex, but the registration accuracy deteriorates and misregistration occurs
Solution Approach 1:
The patent segments the image registration process by performing registration independently for each bone group (cranial bones, facial bones, mandibular bones) rather than attempting to register all bones simultaneously. This segmentation allows each bone group to be registered according to its specific motion characteristics, improving registration accuracy while managing complexity through modular processing.
Solution Approach 2:
The patent applies dynamic registration strategies by selecting different registration methods based on the motion characteristics of each bone group. For example, rigid body registration is used for bones with minimal motion, while deformable registration is applied to bones with larger motion ranges, allowing the system to adapt to varying complexity requirements across different anatomical regions.
2Reliability
If individual bone recognition and association is performed for registration, then registration stability deteriorates in cases of bone surgery or missing bones, but the technique requires detailed bone identification
Solution Approach 1:
The patent segments bones into functional groups (cranial, facial, mandibular) based on their motion characteristics and anatomical relationships. This grouping strategy reduces the complexity of individual bone recognition and association while improving registration stability, as the registration process focuses on group-level alignment rather than requiring precise identification and matching of each individual bone across time points.
Solution Approach 2:
The patent creates a universal registration framework that handles multiple scenarios (normal bones, surgically altered bones, missing bones) through the same bone group-based approach. This multi-functional system automatically adapts to different clinical situations without requiring complex case-specific bone recognition and association procedures, thereby improving reliability while managing complexity.
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
An image processing apparatus includes: an obtaining unit configured to obtain at least one of first identification information and second identification information; a classification unit configured to classify, using the at least one piece of identification information, the plurality of first bones and the plurality of second bones into a first bone group including a bone that moves in association with a motion of a first part of the subject and a second bone group including a bone that moves in association with a motion of a second part different from the first part; and a registration unit configured to perform first registration between the plurality of first bones and the plurality of second bones classified into the first bone group and second registration between the plurality of first bones and the plurality of second bones classified into the second bone group.