MRI Blood Vessel Detection via Image Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Magnetic resonance imaging (MRI) systems face challenges in accurately detecting blood vessels due to high signal interference from blood flow, leading to incorrect identification of tissues as blood vessels, which affects the accuracy of contrast medium detection.
Innovation Solution
A blood vessel detecting apparatus and method that classifies images into specific classes based on the body part region they represent, defining a search region within those classes to improve detection accuracy by using a classification map and probability distribution models to focus on regions with high likelihood of blood vessel presence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If the area of Zdephaser in slice gradient magnetic field is increased to diminish signals from blood in aorta, then signal suppression from blood flow is improved, but high-signal blood during cardiac systole still causes incomplete signal diminishment leading to wrong tissue detection
Solution Approach 1:
The patent segments the image analysis process into multiple steps: first producing axial images, then classifying them into different classes, and finally defining search regions based on the classification. This segmentation allows the system to handle different tissue types differently, improving aorta detection accuracy while maintaining blood signal suppression.
Solution Approach 2:
The patent applies local quality by defining search regions specifically within classified image classes where the aorta is likely to be present. Instead of uniformly processing all images, the system tailors the detection approach to local characteristics of different body regions, thereby improving detection precision without increasing overall signal interference.
2Productivity
If a tracker region is defined based on visual inspection of axial images, then contrast medium detection can be performed, but cross sections of other body parts with resembling signal patterns may be wrongly detected as aorta
Solution Approach 1:
The patent performs preliminary classification of axial images into different classes before defining the search region. This preliminary action distinguishes aorta cross-sections from other body parts based on their characteristic features, ensuring that the subsequent tracker region is defined only in appropriate locations and preventing wrong detection of other tissues.
Solution Approach 2:
The system uses feedback from the classification results to adjust the search region definition. By analyzing the characteristics of classified images and using this information to refine where to place the tracker, the system continuously improves its accuracy in identifying the aorta while maintaining efficient contrast medium detection.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the accuracy of blood vessel detection by reducing misidentification and improving the precision of contrast medium tracking within MRI images.
Implementation Method 1
slice gradient magnetic field
Implementation Method 2
Magnetic resonance imaging apparatus
Data Source
AI summary
An MRI apparatus 1 comprising an image producing unit 101 for producing a plurality of axial images D1 to D10 in a plurality of slices defined in a body part to be imaged containing a blood vessel; a classifying unit 102 for classifying the plurality of axial images D1 to D10 into a plurality of classes I to IV based on which a portion of the imaged body part each of the plurality of axial images D1 to D10 represents; and a defining unit 103 for defining a search region for searching for a blood vessel from within an axial image based on within which of the plurality of classes I to IV the axial image falls.