Segmentation of 4D Cardiac MR Images Using a Probabilistic Atlas and the EM Algorithm

Abstract

In this paper an automatic atlas-based segmentation algorithm for 4D cardiac MR images is proposed. The algorithm is based on the 4D extension of the expectation maximisation (EM) algorithm. The EM algorithm uses a 4D probabilistic cardiac atlas to estimate the initial model parameters and to integrate a-priori information into the classification process. The probabilistic cardiac atlas has been constructed from the manual segmentations of 3D cardiac image sequences of 14 subjects. It provides space and time-varying probability maps for the left and right ventricle, the myocardium, and background structures such as the liver, stomach, lungs and skin. In addition to the probabilistic cardiac atlas, the segmentation algorithm incorporates spatial and temporal contextual information by using 4D Markov Random Fields (MRF). Validation against manual segmentations and computation of the correlation between manual and automatic segmentation on 249 3D volumes were calculated. Results show that the procedure can successfully segment the left ventricle (LV) (r=0.95), myocardium (r=0.83) and right ventricle (RV) (r=0.91).

A bi-ventricular cardiac atlas built from 1000+ high resolution MR images of healthy subjects and an analysis of shape and motion

“…Each voxel of the probabilistic atlas stores a vector value, representing the local probability to be a certain structure (e.g. the left ventricle, right ventricle or myocardium) (Lorenzo-Valdés et al, 2004). Sometimes, a simple label map is used, which is a manual segmentation of the template image without any probability information (Zhuang et al, 2010).…”

Cardiac Image Segmentation from Cine Cardiac MRI Using Graph Cuts and Shape Priors

“…The reported results are the average of these 30 dataset groups. Automatic segmentations were obtained using three methods: graph cuts with intensity information alone (GC), our method using shape priors with graph cuts (GCSP); and the method in Lorenzo-Valdes et al [19] (Met1). The automatic segmentations were compared with manual segmentation using DM and HD.…”

“…In Jolly et al [18], a LV blood pool localization approach is proposed which acts as an initialization for LV segmentation. A 4D probabilistic atlas of the heart and 3D intensity template was used in Lorenzo-Valdes et al [19] to localize the LV. Many other methods have been proposed that segment the LV from short-axis images [20][21][22], multiple views [23,24], and using registration information [25,26].…”

“…Some of the works on LV segmentation also show results for RV segmentation. These include image-based approaches like thresholding [28][29][30], pixel classification approaches [31,32], deformable models [33][34][35][36], active appearance models [27,37], and atlas based methods [19,23,38].…”

Automated frame‐by‐frame endocardial border detection from cardiac magnetic resonance images for quantitative assessment of left ventricular function: Validation and clinical feasibility

“…A majority of techniques described in the literature in this context are based on the use of image intensity gradients, and thus may not necessarily be ideally suited for CMR images, because gradients in these images may not be strong enough to allow accurate endocardial border detection and because the algorithms may be dependent on image quality and the specific pulse sequence used for imaging (23). There are different types of algorithms for boundary detection from CMR images, including approaches based on deformable models (24,25), active shape models (26,27), active appearance models (28,29), and expectation maximization method (30). However, these methods usually require extensive manual tracing for building the model database or for the definition of the training set, thus limiting their clinical application.…”