Segmentation method of intravascular ultrasound images of human coronary arteries
Introduction
Intravascular ultrasound (IVUS) is a medical imaging modality that produces a sequence of cross-sectional frames of the vascular wall of arteries as a catheter is pulled-back inside blood vessels. It has become very useful for studying atherosclerotic diseases [1].
Various segmentation techniques have been developed for IVUS images of coronary arteries. Among methods that are based on statistics of the B-mode image, a Maximum A Posteriori (MAP) estimator was derived using Rayleigh statistics of the signal for the lumen contour segmentation [2]. A multi-surface 3D graph search using Rayleigh distributions and Chan-Vese terms is proposed in [3]. In [4], a knowledge-based approach is used to determine which level of gray corresponds statistically to the different regions of interest, i.e., the intima, plaque and lumen, in the context of the arterial wall segmentation. In [5], a non-parametric probabilistic model is integrated into a shape-driven method for the segmentation of the arterial wall. Among other probabilistic segmentation methods of the luminal borders, let us mention [6], [7], [8].
Several other techniques than the ones based on speckle statistics have also been proposed in the past five years. Edge information alone [9] or combined gray level intensity attributes [10], [11] were used. Other segmentation algorithms are based on different textural features [12], [13], [14], [15]. Gray level intensity and textural information were combined to detect the lumen boundary [6]; edge attributes were added for the external elastic membrane (EEM) [16]. In these methods, different frameworks were used to extract the vessel wall boundaries from the different image information, classifiers being the most prevalent recently [6], [13], [15], [16]. Threshold and contour filtering [12] and binary morphological operations [14] were also proposed. Taki et al. [9] used deformable models while [10] combined them to graph search. Finally, a multi-agent segmentation [11] and a 3D parallel segmentation method [17] were investigated. Less recent techniques can be found in Section 4.
The aim of this work was to show that an adaptation of the fast-marching segmentation method developed in [18], [19] for femoral artery IVUS segmentation is also powerful in the context of coronary imaging. Compared to femoral arteries [18], [19], the artery wall of the coronaries presents a much more complex movement (see Fig. 1, left image). Moreover, since the IVUS images were acquired at a higher frequency (40 MHz) than in the case of femoral arteries (20 MHz), the lumen presents more speckle and hence, its appearance is more variable in the present study than in the context of [18], [19] (see Fig. 1, right image). Therefore, we considered an original modification of the speed function proposed in [18], [19]. A textural gradient, defined in terms of the distribution of the gray levels in the different components of the vessel wall, was introduced in this new speed function. Moreover, mixtures of gamma probability density functions (PDFs) were used to model the gray level distribution of the log-compressed and filtered envelop of the IVUS images [20], [21] that could not be modeled with Rayleigh distributions assuming uniform scattering tissues [22], as was used in our previous work. We also included a process that computes adaptive weights to calibrate the two components of the speed function to accommodate for various ranges of values coming from different image features. The vessel wall boundaries were modeled as layered contours that propagate simultaneously under that new speed function, which is based on a combination of complementary contour and region information. The multiple interfaces were propagated in the IVUS series of images after having been initially positioned using approximate manual segmentations on 2 perpendicular longitudinal views (L-views) of the 3D volume, to allow the tracking of the artery wall. This type of initialization is adapted for the segmentation of large pullbacks (several millimeters) where the vessel wall components might change across the sequence due to the heterogeneity of the image, and when adjacent cross-sections are discontinuous due to the beating heart movement for acquisitions that are not gated. This segmentation model handles contour irregularities, partial shadowing due to calcium inclusions within the plaque, side branches adjacent to the main artery to segment, the presence of a stent, injection of contrast agent or dissection as often observed for atherosclerotic coronary plaques.
Section snippets
In vivo data
A total of 38 in vivo IVUS pullbacks from diseased coronary arteries was obtained from a database of Boston Scientific and clinical studies conducted at the Montreal Heart Institute. From these two sources, we segmented 20 sequences acquired with the Boston Scientific “Galaxy II” scanner, and 18 sequences with the “iLab” echograph. Ultrasound transducers at 40 MHz (mounted on catheters) were used in all cases. The IVUS cross-section image size varied between 8.2 mm and 11.4 mm. In what follows, a
Performance metrics of the FMM segmentations
On average over all sequences, the mean point-to-point and Hausdorff distances for the lumen contours were 0.11 ± 0.03 mm (4.0 ± 1.4%) and 0.33 ± 0.07 mm (11.2 ± 2.3%), respectively. Recall that there are 3207 frames in all performance assessments and that values in parenthesis are MD × 100/D and HD × 100/D, respectively. The mean point-to-point and Hausdorff distances for the EEM contours were 0.10 ± 0.03 mm (2.4 ± 0.7%) and 0.31 ± 0.09 mm (7.6 ± 2.0%), respectively. On average, errors of areas were 0.49 ± 0.18 mm2 (9.3 ±
Discussion
As can be seen from Table 5, Table 6, our study is the most important of the literature with [3] as far as the number of images used for the comparison of algorithmic and manual segmentations is concerned (n = 3207).3 Our validation came from
Conclusion
This study showed with success, for the EEM as well as for the lumen, the good performance of the “fast-marching” algorithm based on the proposed region and contours-based speed functions, compared to other methods of the literature. Note that the PDFs were estimated on each sequence based on the initial contours. In this manner, the proposed method is adaptive to the appearance of the tissues due to various artifacts. An interesting avenue to pursue would be the development of a quantitative
Acknowledgments
This work was jointly supported by a grant from the Ministère du Développement Économique, Innovation et Exportation, Québec, Canada; and Boston Scientific, Freemont, CA, USA. Financial supports were also provided by the Natural Sciences and Engineering Research Council of Canada (grant #138570 - 06). Authors are acknowledging the contributions of Joanne Vincent, Colombe Roy and Ginette Grenier of the IVUS core laboratory of the Montreal Heart Institute for manually segmenting IVUS sequences,
References (52)
- et al.
Automatic segmentation and 3D reconstruction of intravascular ultrasound images for a fast preliminar evaluation of vessel pathologies
Comput Med Imaging Graph
(2007) - et al.
Image analysis techniques for automated IVUS contour detection
Ultrasound Med Biol
(2008) - et al.
Automatic coronary wall segmentation in intravascular ultrasound images using binary morphological reconstruction
Ultrasound Med Biol
(2011) - et al.
Segmentation of the luminal border in intravascular ultrasound B-mode images using a probabilistic approach
Med Image Anal
(2013) - et al.
HoliMAb: a holistic approach for media-adventitia border detection in intravascular ultrasound
Med Image Anal
(2012) - et al.
A critical review and uniformized representation of statistical distributions modeling the ultrasound echo envelope
Ultrasound Med Biol
(2010) - et al.
Ultrasound echo envelope analysis using a homodyned k distribution signal model
Ultrason Imaging
(1994) - et al.
Bayesian level sets for image segmentation
J Visual Commun Image Represent
(2002) - et al.
Decompression and speckle detection for ultrasound images using the homodyned k-distribution
Pattern Recogn Lett
(2003) - et al.
Modeling envelope statistics of blood and myocardium for segmentation of echocardiographic images
Ultrasound Med Biol
(2008)
Three-dimensional automatic quantitative analysis of intravascular ultrasound images
Ultrasound Med Biol
Automated contour detection for high-frequency intravascular ultrasound imaging: a technique with blood noise reduction for edge enhancement
Ultrasound Med Biol
Multi-agent segmentation of IVUS images
Pattern Recogn
Morphometric analysis in three-dimensional intracoronary ultrasound: an in vitro and in vivo study using a novel system for the contour detection of lumen and plaque
Am Heart J
Segmentation of 3D intravascular ultrasound images based on a random field model
Ultrasound Med Biol
A novel active contour model for fully automated segmentation of intravascular ultrasound images: in vivo validation in human coronary arteries
Comput Biol Med
Standardized evaluation methodology and reference database for evaluating IVUS image segmentation
Comp Med Imag Graph
Intravascular ultrasound: novel pathophysiological insights and current clinical applications
Circulation
Fully automatic luminal contour segmentation in intracoronary ultrasound imaging – a statistical approach
IEEE Trans Med Imaging
Plaque development, vessel curvature, and wall shear stress in coronary arteries assessed by X-ray angiography and intravascular ultrasound
Med Image Anal
Shape-driven segmentation of the arterial wall in intravascular ultrasound images
IEEE Trans Inf Technol Biomed
ECOC random fields for lumen segmentation in radial artery IVUS sequences
Automatic segmentation of artery wall in coronary IVUS images: a probabilistic approach
Comput Cardiol
A probabilistic segmentation method for the identification of luminal borders in intravascular ultrasound images
Automatic segmentation of calcified plaques and vessel borders in IVUS images
Int J CARDS
Segmentation of intravascular ultrasound images using graph search and a novel cost function
Cited by (42)
Convolutional networks for the segmentation of intravascular ultrasound images: Evaluation on a multicenter dataset
2022, Computer Methods and Programs in BiomedicineVessel membrane segmentation and calcification location in intravascular ultrasound images using a region detector and an effective selection strategy
2020, Computer Methods and Programs in BiomedicineLumen-intima and media-adventitia segmentation in IVUS images using supervised classifications of arterial layers and morphological structures
2019, Computer Methods and Programs in BiomedicineIVUS images segmentation using spatial fuzzy clustering and hierarchical level set evolution
2019, Computers in Biology and MedicineAutomatic IVUS lumen segmentation using a 3D adaptive helix model
2019, Computers in Biology and MedicineCitation Excerpt :Recently, Sun et al. [9] proposed a method based on Layered Optimal Graph Image Segmentation of Multiple Objects and Surfaces (LOGISMOS) with computer-aided refinement for the improvement of the segmentation result. Using the framework of IVUS segmentation based on active contours or snakes, several methods using bi-dimensional parametric, geometric, geodesic, and region-based active contours (fast-marching method) have been developed [10–13]. An extension to 3D active contour methods based on local properties of the image gradient and image intensity have also been developed to successfully extract contours in IVUS sequences [14–16].