Atherosclerotic cerebrovascular disease leads to formation of lipid-laden plaques that can form emboli when ruptured causing blockage to cerebral vessels. The clinical manifestation of this event sequence is stroke; a leading cause of disability and death. In vivo MR imaging provides detailed image of vascular architecture for the carotid artery making it suitable for analysis of morphological features. Assessing the status of carotid arteries that supplies blood to the brain is of primary interest to such investigations. Reproducible quantification of carotid artery dimensions in MR images is essential for plaque analysis. Manual segmentation being the only method presently makes it time consuming and sensitive to inter and intra observer variability. This paper presents a deformable model for lumen and vessel wall segmentation of carotid artery from MR images. The major challenges of carotid artery segmentation are (a) low signal-to-noise ratio, (b) background intensity inhomogeneity and (c) indistinct inner and/or outer vessel wall. We propose a new, effective object-class uncertainty based deformable model with additional features tailored toward this specific application. Object-class uncertainty optimally utilizes MR intensity characteristics of various anatomic entities that enable the snake to avert leakage through fuzzy boundaries. To strengthen the deformable model for this application, some other properties are attributed to it in the form of (1) fully arc-based deformation using a Gaussian model to maximally exploit vessel wall smoothness, (2) construction of a forbidden region for outer-wall segmentation to reduce interferences by prominent lumen features and (3) arc-based landmark for efficient user interaction. The algorithm has been tested upon T1- and PD-weighted images. Measures of lumen area and vessel wall area are computed from segmented data of 10 patient MR images and their accuracy and reproducibility are examined. These results correspond exceptionally well with manual segmentation completed by radiology experts. Reproducibility of the proposed method is estimated for both intra- and inter-operator studies.
Object segmentation is of paramount interest in many medical imaging applications. Among others, "snake"-an "active contour"-is a popular boundary-based segmentation framework where a spline is continuously deformed to lock onto an object boundary. The dynamics of a snake is governed by different internal and external forces. A major limitation of this framework has been the difficulty in using object-intensity driven features into snake dynamics which may prevent uncontrolled expansion/contraction once the snake leaks through a weak boundary region. In this paper, object-intensity force is effectively introduced into the snake-model using class uncertainty theory. Given a priori knowledge of object/background intensity distributions, class uncertainty yields object/background class of any location and establishes the confidence level of the classification. This confidence level has previously been demonstrated to be high inside the object/background regions and low near boundaries with intermediate intensities. This class uncertainty information adds an expanding (outward) force at locations pertaining to intensity-based object class and a squeezing (inward) force inside background regions. Consequently, the method possesses potential to resist an uncontrolled expansion of the snake (for an expanding type) into the background through a weak boundary while reducing the effect of this force near the boundary using the confidence information. The theory of object class uncertainty induced snake is developed and an implementation with efficient graphical interface is achieved. Preliminary results of application of the proposed snake approach on different images are presented and comparisons with conventional snake approaches are demonstrated.