Graph-Based Motion-Driven Segmentation of the Carotid Atherosclerotique Plaque in 2D Ultrasound Sequences
Carotid plaque segmentation in ultrasound images is a crucial step for carotid atherosclerosis. However, image quality, important deformations and lack of texture are prohibiting factors towards manual or accurate carotid segmentation. We propose a novel
- PDF / 1,858,774 Bytes
- 9 Pages / 439.363 x 666.131 pts Page_size
- 13 Downloads / 200 Views
2
1 Biomedical Simulations and Imaging Laboratory, School of Electrical and Computer Engineering, National Technical University of Athens, Athens, Greece [email protected], [email protected] Center for Biomedical Image Computing and Analytics, Department of Radiology, University of Pennsylvania, Philadelphia, USA [email protected] 3 Center for Visual Computing, Department of Applied Mathematics, Ecole Centrale de Paris, Paris, France [email protected]
Abstract. Carotid plaque segmentation in ultrasound images is a crucial step for carotid atherosclerosis. However, image quality, important deformations and lack of texture are prohibiting factors towards manual or accurate carotid segmentation. We propose a novel fully automated methodology to identify the plaque region by exploiting kinematic dependencies between the atherosclerotic and the normal arterial wall. The proposed methodology exploits group-wise image registration towards recovering the deformation field, on which information theory criteria are used to determine dominant motion classes and a map reflecting kinematic dependencies, which is then segmented using Markov random fields. The algorithm was evaluated on 120 cases, for which manually-traced plaque contours by an experienced physician were available. Promising evaluation results showed the enhanced performance of the algorithm in automatically segmenting the plaque region, while future experiments on additional datasets are expected to further elucidate its potential. Keywords: segmentation, carotid plaque, image registration, ultrasound.
1
Introduction
Carotid plaques are buildups of fatty substances, scar tissue and cholesterol deposits in the inner lining of the carotid artery and they are caused by a chronic degenerative arterial disease called atherosclerosis. The direct association of these arterial lesions with stroke events and the limitations of the current risk-stratification practice both underline the need for the development of computer-aided-diagnosis (CAD) systems for carotid plaques [1]. In particular, the development of effective CAD schemes, which are based on B-mode ultrasound (US) image analysis, is considered a grand challenge by the scientific c Springer International Publishing Switzerland 2015 N. Navab et al. (Eds.): MICCAI 2015, Part III, LNCS 9351, pp. 551–559, 2015. DOI: 10.1007/978-3-319-24574-4_66
552
A. Gastounioti et al.
community, due to the fact that this modality holds a prominent position in screening and follow-up exams of patients with carotid atherosclerosis. In this line of work, and given that manual tracing of regions of interest is always a timeconsuming and strenuous task, automated plaque segmentation in B-mode US is a critical step towards user-independent measurements of novel biomarkers: (a) total plaque area [2], (b) plaque surface irregularities [3], (c) plaque deformations during the cardiac cycle [4], and (d) the underlying material composition of the plaque (texture descriptors) [1]. An attempt to provide a short overview of related
Data Loading...
