Blood Vessel Segmentation Using Surrounding Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging techniques face challenges in accurately segmenting and detecting blood vessels, particularly bifurcations, due to the similarity in intensities between vessels and bones, leading to blurred edges and inefficient segmentation methods.
Innovation Solution
A method involving classification of the vessel surroundings into two classes based on bone structure concentration, using a KNN classifier to select between learning-based ray casting and morphological active contour without edges (MACWE) algorithms for segmentation, ensuring reliable detection of blood vessel bifurcations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single segmentation method is used for all vessels, then the device complexity is reduced, but the segmentation precision deteriorates due to similar intensities between vessels and bones causing blurred edges
Solution Approach 1:
The patent applies different segmentation algorithms to different vessel regions based on their local characteristics. Specifically, it classifies vessel surroundings into bone-proximity and non-bone-proximity regions, then applies ray-casting algorithm to bone-proximity vessels and active contour algorithm to non-bone-proximity vessels. This local differentiation resolves the contradiction by improving segmentation precision for bone-proximity vessels without unnecessarily complicating the entire system.
Solution Approach 2:
The patent changes the segmentation approach based on the parameter of surrounding bone structure concentration. By using a classifier to detect bone proximity and switching algorithms accordingly, the system adapts its segmentation parameters to match the local image characteristics, thereby improving precision without requiring a completely complex system redesign.
2Manufacturing precision
If different segmentation methods are applied based on vessel surrounding classification, then the segmentation precision is improved, but the device complexity increases due to multiple algorithms and classification steps
Solution Approach 1:
The patent segments the overall segmentation task into distinct phases: first classifying vessel surroundings using a simple classifier, then applying different segmentation algorithms to different classes. This task segmentation allows the system to achieve high precision through specialized algorithms while keeping each individual component relatively simple and manageable.
Solution Approach 2:
The patent introduces a classifier as an intermediary component between image input and segmentation algorithms. This classifier acts as a mediator that analyzes vessel surroundings and directs the appropriate segmentation algorithm to each region, thereby coordinating the complexity of multiple algorithms without requiring them to interact directly, which simplifies the overall system architecture.
3Ease of operation
If traditional segmentation methods are used without surrounding classification, then the ease of operation is maintained, but the measurement precision deteriorates due to inability to distinguish vessels from bones in blurred regions
Solution Approach 1:
The patent performs preliminary classification of vessel surroundings before applying segmentation algorithms. By pre-identifying bone-proximity regions using a simple classifier, the system prepares the appropriate segmentation approach in advance, maintaining operational simplicity while ensuring precision is not compromised in challenging regions.
Data Source
AI summary
A method, and an according apparatus and system, for analyzing medical images of blood vessels includes the steps of a) classifying a surrounding of at least one vessel represented in at least one medical image by applying a first classifier to the medical image such that the surrounding of the vessel is assigned to one of at least two surrounding classes, and b) segmentation of the at least one vessel dependent on the surrounding class to which the surrounding of the vessel has been assigned. The invention allows for a reliable segmentation and/or shape detection, in particular bifurcation detection, of blood vessels represented in medical images.