Your browser doesn't support javascript.
loading
Montrer: 20 | 50 | 100
Résultats 1 - 4 de 4
Filtrer
1.
Article de Chinois | WPRIM | ID: wpr-306602

RÉSUMÉ

This paper presents a novel matching method of vessels in 3D reconstruction of heart vessel. The directed binary tree was used to describe the topological structure of heart vessel tree's skeleton. Based on topological property and epipolar property, the branch-points and end-points of each branch level could be automatically matched along the direction of blood stream. Thereupon it is easy to couple the corresponding vessels segments of two angiograms projected in different directions. The 3D heart vessel tree was successfully reconstructed from clinic coronary angiograms, which validates the presented method.


Sujet(s)
Humains , Algorithmes , Simulation numérique , Coronarographie , Méthodes , Maladie coronarienne , Diagnostic , Imagerie diagnostique , Vaisseaux coronaires , Anatomopathologie , Traitement d'image par ordinateur , Imagerie tridimensionnelle , Méthodes , Reconnaissance automatique des formes , Méthodes
2.
Article de Chinois | WPRIM | ID: wpr-357717

RÉSUMÉ

Topological description and matching are committed steps for three-dimensional reconstruction of vessel tree from two angiograms. Binary tree is proposed to describe the two-dimensional vessel tree. "Node Weight" and "Similar Node" are defined in order to get better description. Vessel segments in two images are matched by the preorder traversal of the binary trees, and the method is proved fast and accurate.


Sujet(s)
Humains , Algorithmes , Angiographie , Méthodes , Simulation numérique , Traitement d'image par ordinateur , Méthodes , Imagerie tridimensionnelle , Méthodes , Modèles biologiques , Reconnaissance automatique des formes , Méthodes , Amélioration d'image radiographique , Méthodes
3.
Article de Chinois | WPRIM | ID: wpr-249584

RÉSUMÉ

This paper presents a method for estimating three-dimensional (3D) motion of coronary arteries from single-plane X-ray angiogram sequences on two views. Firstly, original images are preprocessed and two-dimensional (2D) vessel skeletons are extracted. 2D motion estimation is performed along the skeletons in two images. Then geometrical transformation matrix between views is obtained based on perspective projection model for X-ray angiography system, and 3D coordinate of spatial points are calculated. The 3D motion estimation and reconstruction algorithm is applied along the two image sequences to accomplish 3D reconstruction of vessel skeletons and motion vectors between consecutive time instants. Its effectiveness has been demonstrated on clinical single-plane coronary artery angiograms and potential errors are discussed.


Sujet(s)
Humains , Algorithmes , Coronarographie , Méthodes , Maladie coronarienne , Imagerie diagnostique , Électrocardiographie , Imagerie tridimensionnelle , Méthodes , Mouvement , Interprétation d'images radiographiques assistée par ordinateur
4.
Article de Chinois | WPRIM | ID: wpr-354220

RÉSUMÉ

This paper presents an approach for estimating the non-rigid motion of coronary arteries using digital angiographic images. Displacement vectors of vessel points are obtained by finding points correspondences in two successive frames. Smoothness of motion field and vessel deformation measurement are considered in matching, and unmatched regions are also dealt with in the fact that the vessel after deformation may have different size with its original station. The search strategy for optimal matching is carried out using dynamic programming (DP) so that computing cost is reduced. Results of this motion estimation method applied to synthetic and clinical images have shown that the method is accurate, with a root mean square error about one pixel for simulated data. For the case of actual X-ray coronary angiographic images, visual inspection of the detected pairs of points shows that the results are very encouraging.


Sujet(s)
Humains , Angiographie de soustraction digitale , Méthodes , Coronarographie , Méthodes , Vaisseaux coronaires , Physiologie , Mouvement , Interprétation d'images radiographiques assistée par ordinateur
SÉLECTION CITATIONS
DÉTAIL DE RECHERCHE