Contour Arc Fitting for Medical Object Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging algorithms, such as those for identifying bullae in lung parenchyma, fail to differentiate between objects of interest and artifacts, leading to reduced accuracy and confidence in bullae index calculations.
Innovation Solution
An apparatus and method that utilize contour point subsets and fitted arc segments to determine the likelihood of a candidate object belonging to a specific object class by selecting seed elements, generating contour point subsets, fitting arc segments, and assigning seed scores based on selection criteria, allowing differentiation between objects of interest and artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Difficulty of detecting and measuring
If existing medical imaging algorithms identify objects based on size and physical properties, then the detection coverage is improved, but the differentiation between objects of interest and artifacts deteriorates
Solution Approach 1:
The contour of the candidate object is divided into multiple contour point subsets, and each subset is independently fitted with an arc segment. This segmentation allows the algorithm to evaluate different portions of the contour separately, improving the ability to detect subtle shape differences between bullae and artifacts while maintaining comprehensive detection coverage.
Solution Approach 2:
The invention introduces a seed score parameter that quantifies the likelihood of a candidate object being a true bulla based on arc segment fitting quality. By changing from binary classification to probabilistic scoring, the system can differentiate between objects of interest and artifacts with varying degrees of confidence, resolving the contradiction between detection coverage and differentiation accuracy.
2Device complexity
If simple size-based classification is used, then the computational complexity is reduced, but the reliability of object identification deteriorates
Solution Approach 1:
The algorithm performs arc segment fitting on selected contour point subsets rather than the entire contour, and only for candidate objects that meet preliminary size criteria. This partial action approach maintains computational efficiency while introducing reliable shape-based differentiation, resolving the contradiction between computational complexity and identification reliability.
3Measurement precision
If arc segment fitting is applied to all contour points, then the shape analysis precision is improved, but the processing time increases
Solution Approach 1:
The contour is segmented into multiple point subsets, and arc segment fitting is applied to each subset independently. This segmentation enables parallel processing of different contour portions, maintaining high shape analysis precision while reducing overall processing time through distributed computation.
Solution Approach 2:
The algorithm performs preliminary size-based filtering before applying the computationally intensive arc segment fitting. By eliminating obviously ineligible candidates in advance, the system reduces the number of objects requiring detailed shape analysis, thereby reducing processing time while maintaining precision for relevant candidates.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The invention relates to an apparatus for identifying a candidate object in image data and determining a likelihood that the candidate object is an object from an object class. The apparatus comprises an image data receiving unit for receiving image data of an object of the object class, a seed element selecting unit for selecting a portion of the image elements as seed elements, a contour point identifying unit for identifying, for each seed element (SE), contour points, the contour points of a seed element circumscribing a candidate object which comprises the seed element, and a seed score determining unit for determining, for each seed element, a seed score indicative of a likelihood that the candidate object is an object from the object class. The invention allows differentiation between an object of an object class of interest and artifacts.