Intravascular ultrasound image segmentation: a three-dimensional fast-marching method based on gray level distributions

Abstract

Intravascular ultrasound (IVUS) is a catheter based medical imaging technique particularly useful for studying atherosclerotic disease. It produces cross-sectional images of blood vessels that provide quantitative assessment of the vascular wall, information about the nature of atherosclerotic lesions as well as plaque shape and size. Automatic processing of large IVUS data sets represents an important challenge due to ultrasound speckle, catheter artifacts or calcification shadows. A new three-dimensional (3-D) IVUS segmentation model, that is based on the fast-marching method and uses gray level probability density functions (PDFs) of the vessel wall structures, was developed. The gray level distribution of the whole IVUS pullback was modeled with a mixture of Rayleigh PDFs. With multiple interface fast-marching segmentation, the lumen, intima plus plaque structure, and media layers of the vessel wall were computed simultaneously. The PDF-based fast-marching was applied to 9 in vivo IVUS pullbacks of superficial femoral arteries and to a simulated IVUS pullback. Accurate results were obtained on simulated data with average point to point distances between detected vessel wall borders and ground truth <0.072 mm. On in vivo IVUS, a good overall performance was obtained with average distance between segmentation results and manually traced contours <0.16 mm. Moreover, the worst point to point variation between detected and manually traced contours stayed low with Hausdorff distances <0.40 mm, indicating a good performance in regions lacking information or containing artifacts. In conclusion, segmentation results demonstrated the potential of gray level PDF and fast-marching methods in 3-D IVUS image processing.

Segmentation of the luminal border in intravascular ultrasound B-mode images using a probabilistic approach

“…Many of the early approaches were based on the use of local properties of the image such as pixel intensity and gradient information (edges) combined with computational methods including graph search (Sonka et al (1995), von Birgelen et al (1996, Zhang et al (1998)), active surfaces (Klingensmith et al (2000)), active contours (Kovalski et al (2000)), and neural networks (Plissiti et al (2004)). In later approaches, segmentation was accomplished by the use of region and global information including texture (Mojsilovic et al (1997)), gray level variances (Haas et al (2000), Luo et al (2003)) contrast between the regions (Hui- Zhu et al (2002)), statistical properties of the image modeled by Rayleigh distributions using 2D (Haas et al (2000), Brusseau et al (2004)) and 3D information (Cardinal et al (2006)), and by mathematical morphology techniques (dos Santos Filho et al (2006)). …”

“…Wennogle et. al (Wennogle and Hoff (2009)) proposed improvements over the method presented in (Cardinal et al (2006)) which included a preprocessing step to remove motion artifacts, a new directional gradient velocity term, and a post-processing level-set method. Cardinal et al (Cardinal et al (2010)) presented a multiple interface 3D fast-marching method that was based on a combination of gray level probability density functions and the intensity gradient.…”

Texture Analysis and Radial Basis Function Approximation for IVUS Image Segmentation

“…A fuzzy clustering algorithm for adaptive segmentation in IVUS images [8] is investigated by Filho et al Giannoglou et al propose in [9] an automated segmentation method based on a variant of the active contour model. Cardinal et al present a 3D IVUS segmentation where Rayleigh probability density functions (PDFs) are applied for modeling the pixel gray value distribution of the vessel wall structures [10]. An automated approach based on deformable models has been reported by Plissiti et al [11], who employed a Hopfield neural network for the modification and minimization of an energy function as well as a priori vessel geometry knowledge.…”

Probabilistic Segmentation of the Lumen from Intravascular Ultrasound Radio Frequency Data

“…In these images, the lumen presents some texture due to speckle, and lower contrast with respect to the vessel wall tissues. For these images, edge information is not sufficient and therefore, later approaches incorporated prior knowledge using region and global information such as texture [1], gray level variances [2,3], statistical properties of the intensities [4], temporal information (3D segmentation) [5], and discrete wavelet decomposition [6]. Most recent approaches include the use of nonparametric probability densities with global measurements [7], multilevel discrete wavelet frame decomposition [8], discrete wavelet packet transform [9], machine learning classification methods [10], a combination of gray level probability density functions and the intensity gradient [11], linear-filtered gradient vector flow which drives the deformation of a balloon snake [12], and binary morphological object reconstruction [13].…”