Automated segmentation and analysis of vascular structures in magnetic resonance angiographic images

Abstract

The accurate assessment of the presence and extent of vascular disease, and planning of vascular interventions based on MRA requires the determination of vessel dimensions. The current standard is based on measuring vessel diameters on maximum intensity projections (MIPs) using calipers. In order to increase the accuracy and reproducibility of the method, automated analysis of the 3D MR data is required. A novel method for automatically determining the trajectory of the vessel of interest, the luminal boundaries, and subsequent the vessel dimensions is presented. The automated segmentation in 3D uses deformable models, combined with knowledge of the acquisition protocol. The trajectory determination was tested on 20 in vivo studies of the abdomen and legs. In 93% the detected trajectory followed the vessel. Determination of the vessel morphology along a vessel segment is important in grading the presence and extent of possible vascular stenoses. Until recently, most examinations of vascular stenoses were carried out using x-ray angiography (XA). This technology has been regarded as the gold standard in the evaluation of stenoses. However, several problems are associated with this imaging modality. First, since XA is a projection technique, overprojection of vessels can occur even if the view angle is set optimally. Second, an ionizing nephrotoxic contrast agent has to be administered by means of a catheter, a technique that is associated with a definite (although relatively small) morbidity and mortality risk. Finally, the patient and the personnel of the catheterization laboratory are exposed to x-ray radiation.Magnetic resonance angiography (MRA), on the other hand, is a technique that produces three-dimensional (3D) images. Consequently, there is no risk of overprojection of vessels. MRA can be applied without the use of a contrast agent, although the signal-to-noise ratio (SNR) increases if a contrast agent is used. This contrast agent is normally administered intravenously, which presents minimal risks to the patient (1,2).MRA data sets are generally evaluated on 2D maximum intensity projections (MIP); however, it is known that this leads to underestimation of the vessel width, and a decreased SNR. The interpretation is carried out by either visual inspection or by caliper measurements (3-7).To improve the conventional analysis of MRA, it would be desirable to obtain quantitative morphological information directly from the 3D images and not from the projections. To accomplish this, accurate 3D segmentation tools are required.Vessel segmentation of 3D images has been investigated by many researchers (8 -16). However, the majority of this research focused on enhancing the 3D visualization of the vascular structures in the image, and not on accurate quantification of these structures.In this work a novel approach for quantitative vessel analysis of MRA images is introduced and validated. The approach uses knowledge about the image acquisition procedure to accurately determine the vessel boundaries. The techniq...

Effects of Sildenafil on Cerebrovascular Reactivity in Patients with Becker Muscular Dystrophy

“…Dimensions of the MCA were assessed using the stand-alone LKEB-MRA vessel wall analysis software [29]. A stretch of the MCA was chosen that started at its origin at the internal carotid artery and finished approximately 20 mm downstream from that point.…”

High-Resolution 3T MR Angiography of the Carotid Arteries: Comparison of Manual and Semiautomated Quantification of Stenosis

“…[15][16][17][18][19][20][21][22][23][24] Besides data-acquisition techniques, image postprocessing influences the accuracy of the results. A number of postprocessing approaches have been proposed, [25][26][27][28][29][30][31] but few published data address the perfor- mance of commercially available vessel-analysis tools for the assessment of CA stenosis. 32,33 In our study, each tool provided transverse MPR views perpendicular to and along the vessel center line.…”

Towards quantitative analysis of coronary CTA

“…Existing methods for the determination of vessel centerlines can be roughly grouped into two categories: an optimal path calculation based upon the minimization of some cost function [18]; and methods based upon the skeletonization of a segmented vessel [19]. The first type is typically computationally faster, and avoids the error-sensitive segmentation step.…”