Intravascular Ultrasound Lumen Edge Reconstruction via Centerline Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current intravascular ultrasound (IVUS) probes struggle with blood spots in cross-sectional images, requiring subjective doctor interpretation for vascular lumen edge identification, and cannot accurately capture blood vessel curvatures or three-dimensional lesions.
Innovation Solution
A method involving image processing to binarize IVUS images, delineate lumen edges, construct a centerline, and reconstruct blood vessels using a computer device and storage medium, employing algorithms like KMeans and interpolation to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional IVUS probes are used to acquire intravascular cross-sectional images, then vascular tissue information can be obtained, but blood spots appear in the images making lumen edge identification dependent on doctor experience and reducing measurement precision
Solution Approach 1:
The patent replaces the mechanical/subjective interpretation process with an automated image processing system. The system applies binarization, morphological operations, and edge detection algorithms to automatically identify lumen edges, eliminating reliance on doctor experience and subjective interpretation while improving measurement precision and consistency
Solution Approach 2:
The patent enables the image processing system to automatically identify and correct issues without human intervention. The automated pipeline performs thresholding, noise removal, edge detection, and 3D reconstruction tasks independently, making the system self-sufficient in generating accurate vascular measurements without requiring doctor interpretation
2Measurement precision
If current IVUS probes scan tissue information in vascular lumens and around blood vessel walls, then cross-sectional images can be acquired, but the curvatures of tested blood vessels cannot be directly and accurately obtained
Solution Approach 1:
The patent transforms two-dimensional cross-sectional images into a three-dimensional vascular model by detecting the catheter centerline position in each frame and using these positions to reconstruct the 3D vascular structure. This dimensional transformation enables accurate measurement of blood vessel curvatures and preservation of three-dimensional lesion information that cannot be obtained from 2D images alone
Solution Approach 2:
The patent accounts for the asymmetric relationship between the catheter position and the blood vessel geometry. By detecting the centerline position of the catheter in each cross-sectional frame and using these asymmetric position data points, the system reconstructs the true three-dimensional curvature of the blood vessel, rather than assuming symmetric or straight vessel geometry
3Measurement precision
If mainstream IVUS systems present vascular tissue information based on two-dimensional tomography sequences, then images can be displayed, but three-dimensional lesions appear small or of low pixel values making it difficult to identify suspicious information
Solution Approach 1:
The patent converts 2D cross-sectional images into a 3D vascular model that preserves and enhances three-dimensional lesion information. By reconstructing the vascular geometry in 3D space using detected lumen edges and centerline positions, the system makes lesions more visible and measurable, eliminating the problem of lesions appearing small or having low pixel values in 2D presentations
Data Source
Figure 1
Figure 2
Figure 3~4(e)
AI summary
The present application relates to a blood vessel reconstruction method and apparatus, a computer device and a readable storage medium. The method includes: obtaining a plurality of intravascular ultrasound images of a target object (S202); binarizing each of the intravascular ultrasound images to obtain corresponding binary images (S204); traversing pixels of each of the binary images to select the pixels that meet a preset condition as target pixels, and delineating lumen edges of a blood vessel based on the target pixels to obtain segmented regions of interest (S206); obtaining coordinates of a center point of each of the regions of interest based on the segmented regions of interest (S208); constructing a centerline of the blood vessel based on all the obtained coordinates of the center points (S210); and reconstructing the blood vessel by arranging each of the intravascular ultrasound images along the centerline in a preset order (S212). The present application can pick out the pixels with more distinct lumen edge features, improving delineating accuracy, ensuring that the reconstructed three-dimensional model of the blood vessel is close to its actual condition, thus providing doctors with more effective information about the blood vessel.