Conference Paper

Automatic segmentation of pulmonary arterial tree in computer tomography images: Segmentación automática del árbol arterial pulmonar en imágenes TAC

DOI: 10.1109/COLOMCC.2011.5936290 Conference: 6th Colombian Computing Congress (CCC)

ABSTRACT Arterial tree segmentation is a major step for pulmonary embolism analysis and detection. The present document proposes a technique applied to CT images already diagnosed as embolism positive. In order to perform the segmentation of the arterial tree, the lung's region is transformed into a mask, so that the processed volume gets reduced. Then, through thresholding, region growing, and morphological operations the arterial tree is segmented in 3D. The proposed technique was evaluated by means of arterial counting in each segmented slice as well as in the original ones. That technique got a 93.72% of arterial matching on six datasets of CT images.

0 Bookmarks
 · 
122 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Although Pulmonary Embolism (PE) is one of the most common causes of unexpected death in the U.S., it may also be one of the most preventable. Images acquired from 16-slice Computed Tomography (CT) machines of contrast-injected patients provide sufficient resolution for the localization and analysis of emboli located in segmental and sub-segmental arteries. After a PE is found, it is difficult to assess the local characteristics of the affected arterial tree without automation. We propose a method to compute characteristics of the local arterial tree given the location of a PE. The computed information localizes the portion of the arterial tree that is affected by the embolism. Our method is based on the segmentation of the arteries and veins followed by a localized tree computation at the given site. The method determines bifurcation points and the remaining arterial tree. A preliminary segmentation method is also demonstrated to locally eliminate over-segmentation of the arterial tree. The final result can then be used assess the affected lung volume and arterial supply. Initial tests revealed a good ability to compute local tree characteristics of selected sites.
    Proc SPIE 05/2004;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: ABSTRACT Blood vessel segmentation in volumetric data is a necessary prerequisite in various medical imaging applications. Specifically, when considering the application of automatic lung nodule detection in thoracic CT scans, segmented blood vessels can be used in order to resolve local ambiguities based on global considerations and so improve the performance of automated lung nodule detection algorithms. In this paper, a novel regulated morphology approach to fuzzy shape analysis is described in the context of blood vessel extraction in thoracic CT scans. Such a representation is necessary due to noise present in the data and due to the discrete nature of the volumetric data produced by CT scans, particularly the interslice spacing. Regulated morphological operations are a generalization of ordinary morphological operations which relax the extreme strictness inherent in ordinary morphological operations. Based on constraints of collinearity, size, and global direction, a tracking algorithm produces a set of connected trees representing blood vessels and nodules in the volume. The produced tree structures are composed of fuzzy spheres in which the degree of object membership is proportional to the ratio between the occupied volume and the volume of the discrete sphere encompassing it. The proposed algorithm is capable of handling bifurcations and discontinuities of blood vessels. The performance of the blood vessel extraction algorithm described in the paper is evaluated based on a distance measure between a known blood vessel structure and a reconstructed one. As the generation of synthetic data for which the true vessel network is known may not be sufficiently realistic, our evaluation is based on different versions of actual clinical data corrupted by multiplicative Gaussian noise. Preliminary results show that by using the extracted blood vessels it is possible to eliminate approximately 30% of the false positives obtained in previous work., Keywords: medical imaging, computer vision, image processing, mathematical morphology, regulated mor-
    Proc SPIE 05/2004;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Pulmonary vascular tree segmentation has numerous applications in medical imaging and computer-aided diag- nosis (CAD), including detection and visualization of pulmonary emboli (PE), improved lung nodule detection, and quantitative vessel analysis. We present a novel approach to pulmonary vessel segmentation based on a fuzzy segmentation concept, combining the strengths of both threshold and seed point based methods. The lungs of the original image are first segmented and a threshold-based approach identifies core vessel components with a high specificity. These components are then used to automatically identify reliable seed points for a fuzzy seed point based segmentation method, namely fuzzy connectedness. The output of the method consists of the probability of each voxel belonging to the vascular tree. Hence, our method provides the possibility to adjust the sensitivity/specificity of the segmentation result a posteriori according to application-specific requirements, through definition of a minimum vessel-probability required to classify a voxel as belonging to the vascular tree. The method has been evaluated on contrast-enhanced thoracic CT scans from clinical PE cases and demon- strates overall promising results. For quantitative validation we compare the segmentation results to randomly selected, semi-automatically segmented sub-volumes and present the resulting receiver operating characteristic (ROC) curves. Although we focus on contrast enhanced chest CT data, the method can be generalized to other regions of the body as well as to dierent imaging modalities.
    Proc SPIE 04/2008;

Full-text

View
59 Downloads
Available from
Jun 5, 2014