Segmentation Of Cardiac MR Images Research Articles

Segmentation of abdominal CT and cardiac MR images with multi scale visual attention

Segmentation of abdominal CT and cardiac MR images with multi scale visual attention

Hierarchical volumetric transformer with comprehensive attention for medical image segmentation.

Transformer is widely used in medical image segmentation tasks due to its powerful ability to model global dependencies. However, most of the existing transformer-based methods are two-dimensional networks, which are only suitable for processing two-dimensional slices and ignore the linguistic association between different slices of the original volume image blocks. To solve this problem, we propose a novel segmentation framework by deeply exploring the respective characteristic of convolution, comprehensive attention mechanism, and transformer, and assembling them hierarchically to fully exploit their complementary advantages. Specifically, we first propose a novel volumetric transformer block to help extract features serially in the encoder and restore the feature map resolution to the original level in parallel in the decoder. It can not only obtain the information of the plane, but also make full use of the correlation information between different slices. Then the local multi-channel attention block is proposed to adaptively enhance the effective features of the encoder branch at the channel level, while suppressing the invalid features. Finally, the global multi-scale attention block with deep supervision is introduced to adaptively extract valid information at different scale levels while filtering out useless information. Extensive experiments demonstrate that our proposed method achieves promising performance on multi-organ CT and cardiac MR image segmentation.

Myocardial Pathology Segmentation of Multi-modal Cardiac MR Images with a Simple but Efficient Siamese U-shaped Network

Segmentation of multi-modal myocardial pathology images is a challenging task, due to factors such as the heterogeneity caused by large inter-modality and intra-modality intensity variations in multi-modal images, and the diversity of location, shape, and scale of lesion regions. Existing methods based on multi-modal segmentation cannot effectively integrate and utilize complementary information between multiple modalities, leading to the difficulty in segmenting edema and discontinuous scars. In this paper, we propose a simple but efficient U-shaped network, named Siamese U-Net, to solve these problems. There are two aspects to our method. First, we adopt a multi-modal complementary information exploration network (MCIE-Net) to explore the correlations across multi-modal images and simultaneously segment cardiac structures and myocardial pathology. This method is able to fully leverage complementary information between different modalities. Second, to obtain accurate and continuous segmentation of edema and scars, we use a lesion refinement network (LR-Net) with the same architecture as the MCIE-Net, which extracts lesion features to enhance the fusion of lesion information. We conducted extensive experiments on the MyoPS 2020 and MS-CMRSeg 2019 datasets to demonstrate the effectiveness of our proposed approach. We obtained an average Dice score of 0.734 ± 0.088 for the myocardial edema + scars on the MyoPS 2020 test set, a result which outperformed the state-of-the-art method. These results are a 0.9% improvement over the segmentation results of our previous work, and exceed the results of the winner of the MyoPS 2020 challenge by 0.3%.

An Augmented Deep Learning Network with Noise Suppression Feature for Efficient Segmentation of Magnetic Resonance Images

The segmentation of cardiac MR images requires extensive attention as it needs a high level of care and analysis for the diagnosis of affected part. The advent of deep learning technology has paved the way for efficient, automatic and reliable segmentation of medical images for proper diagnosis. This manuscript proposes an augmented end to end trainable deep learning architecture for efficient segmentation of various domains of medical images. The model incorporates the depth wise separable convolution and group normalization as basic building blocks. Moreover, a noise stifler block is also induced between the encoder and decoder to counter the noise in the medical images. This path helps in delineating the precise boundary contours as the noise often reduces the boundary segmentation capability of the segmentation network. The network trained once produces exceedingly good results for the images of other datasets. An improvement of above (5 ± 0.03) % and (3.5 ± 0.02) % was observed in the Jaccard index and Dice score for cardiac MR images. The results are statistically validated as p < 0.05. The automatic computer investigated approach can help in reducing the burden on the medical system by producing accurate and reliable results. The algorithmic results were clinically verified by the senior radiologists by comparison with the manually segmented images. The training time of the network was about 30% less than U-net.

Cardiac MRI Segmentation With a Dilated CNN Incorporating Domain-Specific Constraints

Semantic segmentation of cardiac MR images is a challenging task due to its importance in medical assessment of heart diseases. Having a detailed localization of specific regions of interest such as Right and Left Ventricular Cavities and Myocardium, doctors can infer important information about the presence of cardiovascular diseases, which are today a major cause of death globally. This paper addresses the problem of semantic segmentation in cardiac MR images using a dilated Convolutional Neural Network. Opting for dilated convolutions allowed us to work in full resolution throughout the network's layers, preserving localization accuracy, while maintaining a relatively small number of trainable parameters. To assist the network's training process we designed a custom loss function. Furthermore, we developed new augmentation techniques and also adapted existing ones, to cope for the lack of sufficient training images. Consequently, the training set increases not only by amount, but by substance as well, and the network trains quickly and efficiently without overfitting. Our pre- and post-processing steps are also crucial to the whole process. We apply our methodology for the Right and Left Ventricles (RV, LV) and also the Myocardium (MYO) according to the Automated Cardiac Diagnosis Challenge (ACDC) with promising results. Submitting our algorithm's predictions to the Post-2017-MICCAI-challenge testing phase, we achieved similar scores (average Dice coefficient 0.916) on the test data set compared to the state of the art featured in the ACDC leaderboard, but with significantly fewer parameters than the leading method. Our approach outperforms other methods featuring dilated convolutions in this challenge up until now.

Improving the Generalizability of Convolutional Neural Network-Based Segmentation on CMR Images.

Background: Convolutional neural network (CNN) based segmentation methods provide an efficient and automated way for clinicians to assess the structure and function of the heart in cardiac MR images. While CNNs can generally perform the segmentation tasks with high accuracy when training and test images come from the same domain (e.g., same scanner or site), their performance often degrades dramatically on images from different scanners or clinical sites.Methods: We propose a simple yet effective way for improving the network generalization ability by carefully designing data normalization and augmentation strategies to accommodate common scenarios in multi-site, multi-scanner clinical imaging data sets. We demonstrate that a neural network trained on a single-site single-scanner dataset from the UK Biobank can be successfully applied to segmenting cardiac MR images across different sites and different scanners without substantial loss of accuracy. Specifically, the method was trained on a large set of 3,975 subjects from the UK Biobank. It was then directly tested on 600 different subjects from the UK Biobank for intra-domain testing and two other sets for cross-domain testing: the ACDC dataset (100 subjects, 1 site, 2 scanners) and the BSCMR-AS dataset (599 subjects, 6 sites, 9 scanners).Results: The proposed method produces promising segmentation results on the UK Biobank test set which are comparable to previously reported values in the literature, while also performing well on cross-domain test sets, achieving a mean Dice metric of 0.90 for the left ventricle, 0.81 for the myocardium, and 0.82 for the right ventricle on the ACDC dataset; and 0.89 for the left ventricle, 0.83 for the myocardium on the BSCMR-AS dataset.Conclusions: The proposed method offers a potential solution to improve CNN-based model generalizability for the cross-scanner and cross-site cardiac MR image segmentation task.

A distance map regularized CNN for cardiac cine MR image segmentation.

Cardiac image segmentation is a critical process for generating personalized models of the heart and for quantifying cardiac performance parameters. Fully automatic segmentation of the left ventricle (LV), the right ventricle (RV), and the myocardium from cardiac cine MR images is challenging due to variability of the normal and abnormal anatomy, as well as the imaging protocols. This study proposes a multi-task learning (MTL)-based regularization of a convolutional neural network (CNN) to obtain accurate segmenation of the cardiac structures from cine MR images. We train a CNN network to perform the main task of semantic segmentation, along with the simultaneous, auxiliary task of pixel-wise distance map regression. The network also predicts uncertainties associated with both tasks, such that their losses are weighted by the inverse of their corresponding uncertainties. As a result, during training, the task featuring a higher uncertainty is weighted less and vice versa. The proposed distance map regularizer is a decoder network added to the bottleneck layer of an existing CNN architecture, facilitating the network to learn robust global features. The regularizer block is removed after training, so that the original number of network parameters does not change. The trained network outputs per-pixel segmentation when a new patient cine MR image is provided as an input. We show that the proposed regularization method improves both binary and multi-class segmentation performance over the corresponding state-of-the-art CNN architectures. The evaluation was conducted on two publicly available cardiac cine MRI datasets, yielding average Dice coefficients of 0.84±0.03 and 0.91±0.04. We also demonstrate improved generalization performance of the distance map regularized network on cross-dataset segmentation, showing as much as 42% improvement in myocardium Dice coefficient from 0.56±0.28 to 0.80±0.14. We have presented a method for accurate segmentation of cardiac structures from cine MR images. Our experiments verify that the proposed method exceeds the segmentation performance of three existing state-of-the-art methods. Furthermore, several cardiac indices that often serve as diagnostic biomarkers, specifically blood pool volume, myocardial mass, and ejection fraction, computed using our method are better correlated with the indices computed from the reference, ground truth segmentation. Hence, the proposed method has the potential to become a non-invasive screening and diagnostic tool for the clinical assessment of various cardiac conditions, as well as a reliable aid for generating patient specific models of the cardiac anatomy for therapy planning, simulation, and guidance.

Cine Cardiac MRI Slice Misalignment Correction Towards Full 3D Left Ventricle Segmentation.

Accurate segmentation of the left ventricle (LV) blood-pool and myocardium is required to compute cardiac function assessment parameters or generate personalized cardiac models for pre-operative planning of minimally invasive therapy. Cardiac Cine Magnetic Resonance Imaging (MRI) is the preferred modality for high resolution cardiac imaging thanks to its capability of imaging the heart throughout the cardiac cycle, while providing tissue contrast superior to other imaging modalities without ionizing radiation. However, there exists an inevitable misalignment between the slices in cine MRI due to the 2D + time acquisition, rendering 3D segmentation methods ineffective. A large part of published work on cardiac MR image segmentation focuses on 2D segmentation methods that yield good results in mid-slices, however with less accurate results for the apical and basal slices. Here, we propose an algorithm to correct for the slice misalignment using a Convolutional Neural Network (CNN)-based regression method, and then perform a 3D graph-cut based segmentation of the LV using atlas shape prior. Our algorithm is able to reduce the median slice misalignment error from 3.13 to 2.07 pixels, and obtain the blood-pool segmentation with an accuracy characterized by a 0.904 mean dice overlap and 0.56 mm mean surface distance with respect to the gold-standard blood-pool segmentation for 9 test cine MR datasets.

Segmentation of cardiac tagged MR images using a snake model based on hybrid gradient vector flow

In the segmentation of cardiac tagging magnetic resonance (tMR) images, it is difficult to segment the left ventricle automatically by using the traditional segmentation model because of the interference caused by the tags. A new snake model based on hybrid gradient vector flow (HGVF) is proposed by us to improve this segmentation. Due to the different characteristics between endocardium and epicardium of the left ventricle (LV), several gradient vector flows (GVFs) with distinctive boundary information would be fused to segment these two sub regions individually. For segmentation of endocardium, we construct a new HGVF in snake model fused by three independent GVFs. These flows are respectively exported from the original cardiac tMR image, the tags-removed image and the local-filtered image. On the other hand, since the epicardium is with a nearly-circle shape, we construct the other HGVF which is composed of two different GVFs. One of them is derived from the tags-removed image either and the other one is derived from the ideal circle-shape image. Some experiments have been done to validate our new segmentation model. The average overlap of the endocardium segmentation is 89.67% (its mean absolute distance is 1.86 pixels), and the average overlap of the epicardium segmentation is 95.88% (its mean absolute distance is 1.64 pixels). Experimental results show that the proposed method improves the segmentation performance compared to some available methods effectively.

A Modified Active Contour Model for Cardiac MR Image Segmentation Based on Entropy Information

In this article, a Modified Geodesic Active Contour Model has been presented. Normally, Active Contour-based methods are categorized as either edge-based models or region-based models. Active Contour Models not only uses data on Gradient, but also can use global features of image for evolutionary curve to move and segment target boundaries which leads to detection of their weak ones. This type of function, introduce a window to calculate the average surrounding data on any pixel, and fashion a Signed Pressure Force based on entropy to make the contour reach the boundaries. In order to optimizing the efficiency and stability of the algorithm, this article has provided two sorts of level set methods for segmentation to be done. Simulation results demonstrate that due process can provide precise readings even in cases of inhomogeneity or not well-defined edges.

Multi-atlas label fusion using hybrid of discriminative and generative classifiers for segmentation of cardiac MR images.

Multi-atlas segmentation first registers each atlas image to the target image and transfers the label of atlas image to the coordinate system of the target image. The transferred labels are then combined, using a label fusion algorithm. In this paper, we propose a novel label fusion method which aggregates discriminative learning and generative modeling for segmentation of cardiac MR images. First, a probabilistic Random Forest classifier is trained as a discriminative model to obtain the prior probability of a label at the given voxel of the target image. Then, a probability distribution of image patches is modeled using Gaussian Mixture Model for each label, providing the likelihood of the voxel belonging to the label. The final label posterior is obtained by combining the classification score and the likelihood score under Bayesian rule. Comparative study performed on MICCAI 2013 SATA Segmentation Challenge demonstrates that our proposed hybrid label fusion algorithm is accurate than other five state-of-the-art label fusion methods. The proposed method obtains dice similarity coefficient of 0.94 and 0.92 in segmenting epicardium and endocardium respectively. Moreover, our label fusion method achieves more accurate segmentation results compared to four other label fusion methods.

A collaborative resource to build consensus for automated left ventricular segmentation of cardiac MR images

A collaborative framework was initiated to establish a community resource of ground truth segmentations from cardiac MRI. Multi-site, multi-vendor cardiac MRI datasets comprising 95 patients (73 men, 22 women; mean age 62.73±11.24years) with coronary artery disease and prior myocardial infarction, were randomly selected from data made available by the Cardiac Atlas Project (Fonseca et al., 2011). Three semi- and two fully-automated raters segmented the left ventricular myocardium from short-axis cardiac MR images as part of a challenge introduced at the STACOM 2011 MICCAI workshop (Suinesiaputra et al., 2012). Consensus myocardium images were generated based on the Expectation-Maximization principle implemented by the STAPLE algorithm (Warfield et al., 2004). The mean sensitivity, specificity, positive predictive and negative predictive values ranged between 0.63 and 0.85, 0.60 and 0.98, 0.56 and 0.94, and 0.83 and 0.92, respectively, against the STAPLE consensus. Spatial and temporal agreement varied in different amounts for each rater. STAPLE produced high quality consensus images if the region of interest was limited to the area of discrepancy between raters. To maintain the quality of the consensus, an objective measure based on the candidate automated rater performance distribution is proposed. The consensus segmentation based on a combination of manual and automated raters were more consistent than any particular rater, even those with manual input. The consensus is expected to improve with the addition of new automated contributions. This resource is open for future contributions, and is available as a test bed for the evaluation of new segmentation algorithms, through the Cardiac Atlas Project (www.cardiacatlas.org).

Segmentation of Short Axis Cardiac MR Images using PCA with Active Appearance Model

We describe the method for segmentation of Left Ventricle (LV) in short axis cardiac MR Images in order to visibly identify the LV, the endocardium and the epicardium. The interpretation and evaluation of images obtained from Magnetic Resonance Imaging (MRI) is not a trivial task and also it is very time- consuming if done completely manually. This is especially the case for higher dimensional image data such as volumetric and/or time-dependent data such as the Cardiac Images. The objective of this study was show clearly the LV in particular so that any deviation from the standard dimensions in terms of shape, size or texture, can be unmistakably identified. Early diagnosis of heart abnormalities can potentially decrease the risk of heart attacks as well as benefit people who have just undergone Heart Transplant Surgery. For segmentation of the Cardiac MRI, Principal Component Analysis (PCA) is used in the Active Appearance Model (AAM) building process. The extraction of statistical features with PCA can be regarded as a projection of high-dimensional data vectors into a space of much lower dimensionality. The process of building a computerized model is equated with making the computer learn possible appearances of an object. The basis for such a learning process is the training set which is prepared from the data obtained from a reputed hospital in Pune. I. Introduction According to the Organ Procurement and Transplantation Network (OPTN), out of approximately 4000 heart transplants performed every year in the United States, 543 patients die in the first year, only 78% survive 3 years after the transplant and 72% after 5 years. Most of these deaths are due to rejection of the transplanted organs. To prevent rejection, patients are given immuno-suppressants, which suppress the immune system to enable acceptance of the newly transplanted organ. However, excess of these drugs destroy the patient's immune system making them susceptible to numerous diseases. These patients are monitored closely by doctors to adjust the dosage of immuno-suppressants given to them. One method of doing this is by biopsy, which involves taking a small piece of the heart muscle and inspecting it under the microscope for damaged cells. This procedure is highly invasive and prone to sampling errors. MRI can be used to effectively monitor any functional abnormality of the heart in a non-invasive manner. Segmentation of the cardiac MRI images into epicardium and ventricles helps to detect any deformities and thus could indicate the on-setting of rejection. This segmentation can be performed manually by experts, but it is a time consuming process, and may delay the detection of rejection at a crucial stage. The manually performed segmentation also suffers the problem of inconsistency. Hence, efforts have been made to make this segmentation procedure completely automatic. (1, 2) The heart is a muscular organ of the circulatory system that constantly pumps blood throughout the body. As seen in figure 1, the heart has four separate chambers. The upper chambers are called the left and right atria, and the lower chambers are called the left and right ventricles. The septum separates the right and left sides of the heart. The endocardium is the inner surface of the myocardium (the heart muscle tissue) and the epicardium is the outer surface of the myocardium. The muscular tissues that attach to the lower portion of the interior wall of the ventricles are the papillary muscles. The periodic motion of the walls of the heart chambers during one heartbeat is referred to as one cardiac cycle. One heartbeat consists of two phases, namely systole and diastole. The rhythmic contraction of the ventricles, by which blood is driven through the aorta and pulmonary artery is the systole. Diastole is the normal rhythmically occurring relaxation and dilatation of the ventricles, during which they fill with blood. The right ventricle delivers blood to the lungs. The left ventricle delivers blood to the rest of the body. So the left ventricular wall is thicker. This fact is often used for distinguishing between the left and right ventricles. Here, we study the MRI images of the transverse section of the heart and segment them into epicardium, right ventricle, and left ventricle. (1, 2) Magnetic Resonance Imaging (MRI) is a method of creating images of the inside of opaque organs in living organisms. When the object to be imaged is placed in a powerful uniform magnetic field, the spins of the hydrogen nuclei with non-zero spin numbers, within the tissue, all align in one of two opposite directions: parallel to the magnetic field or three orthogonal image gradients are applied, to selectively image the different voxels (3-D pixels). They are the slice selection gradient, the phase encoding gradient, and the frequency

Modified active contour model and Random Walk approach for left ventricular cardiac MR image segmentation

AbstractThere is an increased rate of inconsistency in discriminating objects in cardiac magnetic resonance images with increasing disease varieties. This correspondence suggests a combination of two methods to segment the left ventricle in such images. The importance of the weighting function used in Random Walk approach is analyzed. Difference of Laplacian of Gaussian is demonstrated to be efficacious as the desired weighting function in extracting the blood pool boundary. Subsequently, the epicardial wall is sectioned out using an active contour model that considers the blood pool boundary as its initial contour to grow. Finally in order to validate the analysis, the algorithm is tested by comparing these determined borders with the ground truth images. Copyright © 2011 John Wiley &amp; Sons, Ltd.

An Image-Based Comprehensive Approach for Automatic Segmentation of Left Ventricle from Cardiac Short Axis Cine MR Images

Segmentation of the left ventricle is important in the assessment of cardiac functional parameters. Manual segmentation of cardiac cine MR images for acquiring these parameters is time-consuming. Accuracy and automation are the two important criteria in improving cardiac image segmentation methods. In this paper, we present a comprehensive approach to segment the left ventricle from short axis cine cardiac MR images automatically. Our method incorporates a number of image processing and analysis techniques including thresholding, edge detection, mathematical morphology, and image filtering to build an efficient process flow. This process flow makes use of various features in cardiac MR images to achieve high accurate segmentation results. Our method was tested on 45 clinical short axis cine cardiac images and the results are compared with manual delineated ground truth (average perpendicular distance of contours near 2 mm and mean myocardium mass overlapping over 90%). This approach provides cardiac radiologists a practical method for an accurate segmentation of the left ventricle.

Segmentation of cardiac cine MR images of left and right ventricles: Interactive semiautomated methods and manual contouring by two readers with different education and experience

To test interactive semiautomated methods (ISAM) vs. manual contouring (MC) in segmenting cardiac cine MR images. Short-axis images of 10 consecutive patients (1.5-81.5 years of age) were evaluated by a trained radiologist (R1) and a low-trained engineer (R2). Each of them performed four independent reading sessions: two using ISAM and two using MC. Left ventricle (LV) myocardial mass (LVMM), LV ejection fraction (LVEF), and right ventricle (RV) ejection fraction (RVEF) were obtained. Bland-Altman analysis and Wilcoxon test were used. The bias +/- 2 standard deviations (SD) of ISAM vs. MC for LVMM (g) was -5.7 +/- 13.4 (R1) and -5.5 +/- 26.3 (R2); for LVEF (%) it was -1.4 +/- 13.0 and -2.9 +/- and 6.8; for RVEF (%) it was 2.6 +/- 17.0 and 1.0 +/- 16.7. Considering both readers/methods, intraobserver bias +/- 2 SD ranged from 0.3 +/- 25.3 to -6.8 +/- 23.0, from 0.2 +/- 8.0 to -4.4 +/- 15.8, and from -0.0 +/- 26.4 to -4.6 +/- 27.8, respectively. Interobserver bias +/- 2 SD was -25.9 +/- 46.0 (ISAM) and 26.1 +/- 36.4 (MC), -1.4 +/- 8.6 (ISAM) and 0.1 +/- 17.9 (MC), and 0.7 +/- 23.3 and 2.3 +/- 29.8, respectively. Larger SDs were systematically found for RVEF vs. LVEF. Segmentation times: five minutes for LV with ISAM (both readers); for LV with MC, six (R1) vs. nine minutes (R2) (P < 0.001); five to six minutes for RV (both methods /readers). R2 significantly reduced LV segmentation times from nine (MC) to five minutes (ISAM) (P < 0.001). A highly reproducible LV segmentation was performed in a short time by R1. The advantage of ISAM vs. MC for LV segmentation was a time saving only for R2. For RVEF, a lower reproducibility was observed for both methods and readers.

Bidirectional segmentation of three-dimensional cardiac MR images using a subject-specific dynamical model.

Statistical model-based segmentation of the left ventricles has received considerable attention these years. While many statistical models have been shown to improve segmentation results, most of them either belong to (1) static models (SM) that neglect the temporal coherence of a cardiac sequence, or (2) generic dynamical models (GDM) that neglect the individual differences of cardiac motion. In this paper, we propose a subject-specific dynamical model (SSDM) that can simultaneously handle inter-subject variability and temporal cardiac dynamics (intra-subject variability). We also design a dynamic prediction algorithm that can progressively predict the shape of a new cardiac sequence at a given frame based on the shapes observed in earlier frames. Furthermore, to reduce the accumulation of the segmentation errors throughout the entire sequence, we take into account the periodic nature of cardiac motion and perform bidirectional segmentation from a certain frame in a cardiac sequence. "Leave-one-out" validation on 32 sequences show that our algorithm can capture local shape variations and suppress the propagation of segmentation errors.

Comparative evaluation of active contour model extensions for automated cardiac MR image segmentation by regional error assessment

In the field of cardiac MR image segmentation, active contour models, or snakes have been extensively used, owing to their promising results and to the numerous extensions proposed to improve their performance. This paper explores a methodology for evaluating cardiac MR image segmentation algorithms, which assesses the distance between computer-generated and the observer's hand-outlined boundaries. This metric was applied to various external force extensions of the traditional snake, since no systematic comparison has been performed. Cardiac MRI from six patients were analyzed. Imaging was performed on a 1.5 T MR scanner with ECG-gated balanced steady-state free precession (b-SSFP) sequences. Segmentation performances were established for traditional snake, gradient vector flow snake, standard- and guided- pressure force-based snake. The use of a pre-treatment with non-linear anisotropic filtering was also compared to non-filtered images. Agreement between manual and segmentation algorithms was satisfactory for ejection fraction for every segmentation scheme. However end-systolic and end-diastolic volumes were systematically underestimated. The developed regional error metric provided a more rigorous evaluation of the segmentation schemes in comparison to the classical derived parameters based on left ventricle volume estimation, usually used in functional cardiac MR studies. These derived parameters can furthermore mask local segmentation errors.

STACS: new active contour scheme for cardiac MR image segmentation

The paper presents a novel stochastic active contour scheme (STACS) for automatic image segmentation designed to overcome some of the unique challenges in cardiac MR images such as problems with low contrast, papillary muscles, and turbulent blood flow. STACS minimizes an energy functional that combines stochastic region-based and edge-based information with shape priors of the heart and local properties of the contour. The minimization algorithm solves, by the level set method, the Euler-Lagrange equation that describes the contour evolution. STACS includes an annealing schedule that balances dynamically the weight of the different terms in the energy functional. Three particularly attractive features of STACS are: 1) ability to segment images with low texture contrast by modeling stochastically the image textures; 2) robustness to initial contour and noise because of the utilization of both edge and region-based information; 3) ability to segment the heart from the chest wall and the undesired papillary muscles due to inclusion of heart shape priors. Application of STACS to a set of 48 real cardiac MR images shows that it can successfully segment the heart from its surroundings such as the chest wall and the heart structures (the left and right ventricles and the epicardium.) We compare STACS' automatically generated contours with manually-traced contours, or the "gold standard," using both area and edge similarity measures. This assessment demonstrates very good and consistent segmentation performance of STACS.

Automated cardiac MR image segmentation: theory and measurement evaluation

We present a new approach to magnetic resonance image segmentation with a Gradient-Vector-Flow-based snake applied to selective smoothing filtered images. The system also allows automated image segmentation in the presence of grey scale inhomogeneity, as in cardiac Magnetic Resonance imaging. Removal of such inhomogeneities is a difficult task, but we proved that using non-linear anisotropic diffusion filtering, myocardium edges are selectively preserved. The approach allowed medical data to be automatically segmented in order to track not only endocardium, which is usually a less difficult task, but also epicardium in anatomic and perfusion studies with Magnetic Resonance. The method developed proceeds in three distinct phases: (a) an anisotropic diffusion filtering tool is used to reduce grey scale inhomogeneity and to selectively preserve edges; (b) a Gradient-Vector-Flow-based snake is applied on filtered images to allow capturing a snake from a long range and to move into concave boundary regions; and (c) an automatic procedure based on a snake is used to fit both endocardium and epicardium borders in a multiphase, multislice examination. A good agreement ( P<0.001) between manual and automatic data analysis, based on the mean difference±SD, was assessed in a pool of 907 cardiac function and perfusion images.