In this paper, we present an integrated model-based processing scheme for cardiac magnetic resonance imaging (MRI), embedded in an interactive computing environment suitable for quantitative cardiac analysis, which provides a set of functions for the extraction, modeling, and visualization of cardiac shape and deformation. The methods apply four-dimensional (4-D) processing (three spatial and one temporal) to multiphase multislice MRI acquisitions and produce a continuous 4-D model of the myocardial surface deformation. The model is used to measure diagnostically useful parameters, such as wall motion, myocardial thicking, and myocardial mass measurements. The proposed model-based shape extraction method has the advantage of integrating local information into an overall representation and produces a robust description of cardiac cavities. A learning segmentation process that incorporates a generating-shrinking neural network is combined with a spatiotemporal parametric modeling method through functional basis decomposition. A multiscale approach is adopted, which uses at each step a coarse-scale model defined at the previous step in order to constrain the boundary detection. The representation accuracy starts from a coarse but smooth estimation of the approximate cardiac shape and is gradually increased to the desired detail. The main advantages of the proposed methods are efficiency, lack of uncertainty about convergence, and robustness to image artifacts. Experimental results obtained from application to clinical multislice multiphase MRI examinations of normal volunteers and patients with medical record of myocardial infarction were satisfactory in terms of accuracy and robustness.

We propose an image segmentation method based on morphological decomposition and graph-based region merging using contour dynamics. The input image is initially decomposed into a set of primitive homogeneous regions through the morphological watershed transform applied to the image intensity gradient magnitude. This decomposition is represented by a region adjacency graph (RAG) that is input to a hierarchical merging process in which neighboring regions of high similarity are merged. The region similarity criterion is based on the concept of watershed contour dynamics. The robustness of the segmentation to the presence of noise and/or low contrast is improved by a regularization of the contour dynamics. Experimental results on various kinds of synthetic and real images, as well as comparison of the proposed method with other wellknown region merging algorithms are presented.

We propose a novel method for the extraction of coronary artery skeletons (centerlines) by analyzing the topographic features of coronary angiogram images Initially, the angiogram is pre-processed for noise reduction and artery enhancement through directional morphological filtering. The topographic features of the resulting image are detected based on first and second-order image derivatives which characterize the local differential image structure. Using an artery model of a smooth elongated object with an approximately Gaussian smoothed semi-elliptical profile, the candidate skeleton areas are detected as sets of points consisting of ridges, saddle points and peaks. False skeleton areas, produced due to the noise sensitivity of the differentiation filters, have small size and are eliminated by connected component labeling (CCL). CCL may cause the elimination of a few true skeletons which are recovered by the morphological operation of binary reconstruction. Experimental results on clinical coronary angiograms are presented and discussed indicating the robust performance of the proposed method.

A method for extraction and labeling of the coronary arterial tree (CAT) using minimal user supervision in single-view angiograms is proposed. The CAT structural description (skeleton and borders) is produced, along with quantitative information for the artery dimensions and assignment of coded labels, based on a given coronary artery model represented by a graph. The stages of the method are: 1) CAT tracking and detection; 2) artery skeleton and border estimation; 3) feature graph creation; and iv) artery labeling by graph matching. The approximate CAT centerline and borders are extracted by recursive tracking based on circular template analysis. The accurate skeleton and borders of each CAT segment are computed, based on morphological homotopy modification and watershed transform. The approximate centerline and borders are used for constructing the artery segment enclosing area (ASEA), where the defined skeleton and border curves are considered as markers. Using the marked ASEA, an artery gradient image is constructed where all the ASEA pixels (except the skeleton ones) are assigned the gradient magnitude of the original image. The artery gradient image markers are imposed as its unique regional minima by the homotopy modification method, the watershed transform is used for extracting the artery segment borders, and the feature graph is updated. Finally, given the created feature graph and the known model graph, a graph matching algorithm assigns the appropriate labels to the extracted CAT using weighted maximal cliques on the association graph corresponding to the two given graphs. Experimental results using clinical digitized coronary angiograms are presented.

A hybrid multidimensional image segmentation algorithm is proposed, which combines edge and region-based techniques through the morphological algorithm of watersheds. An edge-preserving statistical noise reduction approach is used as a preprocessing stage in order to compute an accurate estimate of the image gradient. Then, an initial partitioning of the image into primitive regions is produced by applying the watershed transform on the image gradient magnitude. This initial segmentation is the input to a computationally efficient hierarchical (bottom-up) region merging process that produces the final segmentation. The latter process uses the region adjacency graph (RAG) representation of the image regions. At each step, the most similar pair of regions is determined (minimum cost RAG edge), the regions are merged and the RAG is updated. Traditionally, the above is implemented by storing all RAG edges in a priority queue. We propose a significantly faster algorithm, which additionally maintains the so-called nearest neighbor graph, due to which the priority queue size and processing time are drastically reduced. The final segmentation provides, due to the RAG, one-pixel wide, closed, and accurately localized contours/surfaces. Experimental results obtained with two-dimensional/three-dimensional (2-D/3-D) magnetic resonance images are presented.

Hierarchical prioritized predictive image coding methods are presented for progressive image transmission. The three main coder stages are hierarchical transform, prioritized coefficient coding and adaptive multiple distribution entropy coding (AMDEC). Following a hierarchical subband decomposition of the original image, the quantized coefficients are scalar quantized and then coded using novel hierarchical partition priority coding (HPPC) and predictive HPPC (PHPPC) algorithms. Given a suitable partitioning of their absolute range, the quantized detail coefficients are ordered based on both their decomposition level and partition and, then, are DPCM coded along with the corresponding address map. The use of space filling scanning further reduces the coding cost. Finally, AMDEC is applied to the HPPC/PHPPC output. Experimental results demonstrate the performance of the proposed compression methods.

Image compression methods for progressive transmission using optimal hierarchical decomposition, partition priority coding (PPC), and multiple distribution entropy coding (MDEC) are presented. In the proposed coder, a hierarchical subband/wavelet decomposition transforms the original image. The analysis filter banks are selected to maximize the reproduction fidelity in each stage of progressive image transmission. An efficient triple-state differential pulse code modulation (DPCM) method is applied to the smoothed subband coefficients, and the corresponding prediction error is Lloyd-Max quantized. Such a quantizer is also designed to fit the characteristics of the detail transform coefficients in each subband, which are then coded using novel hierarchical PPC (HPPC) and predictive HPPC (PHPPC) algorithms. More specifically, given a suitable partitioning of their absolute range, the quantized detail coefficients are ordered based on both their decomposition level and partition and then are coded along with the corresponding address map. Space filling scanning further reduces the coding cost by providing a highly spatially correlated address map of the coefficients in each PPC partition. Finally, adaptive MDEC is applied to both the DPCM and HPPC/PHPPC outputs by considering a division of the source (quantized coefficients) into multiple subsources and adaptive arithmetic coding based on their corresponding histograms. Experimental results demonstrate the great performance of the proposed compression methods.

Motion vector field (MVF) prediction methods are presented followed by a restoration method. These methods combined with a proposed motion compensated (MC) video coding scheme are suitable for low bit rate transmission. An expression is derived for the initial estimate of the working MVF based on the preceding MVF. Spatio-temporally adaptive regularization is applied using neighborhood information. The output MVF is used as the initial prediction estimate for a Kalman MVF restoration approach. By applying this method to both the encoder and decoder, the resulting MC MVF and image intensity temporal updates are coded and transmitted. The restoration method produces accurate estimates of the MVF, thus resulting in a significant transmission cost reduction. Experiments with standard video-conference image sequences demonstrate the improved performance of the proposed scheme.

In this paper, a thorough review is presented of noise reduction filters for digital image sequences. Detailed descriptions of several spatiotemporal and temporal noise reduction algorithms are provided. To aid in comparing between these different algorithms, we classify them based on their support (i.e., 3-D or 1-D filter) and whether or not motion compensation is employed. Several algorithms from each of the four categories are implemented and tested on real sequences degraded to various signal-to-noise ratios. These experimental results are discussed and analyzed to determine the overall advantages and disadvantages of the four general classifications, as well as, the individual filters.>

An approach for estimating the motion of arteries in digital angiographic image sequences is proposed. Binary skeleton images are registered using an elastic registration algorithm in order to estimate the motion of the corresponding arteries. This algorithm operates recursively on the skeleton images by considering an autoregressive (AR) model of the deformation in conjunction with a dynamic programming (DP) algorithm. The AR model is used at the pixel level and provides a suitable cost function to DP through the innovation process. In addition, a moving average (MA) model for the motion of the entire skeleton is used in combination with the local AR model for improved registration results. The performance of this motion estimation method is demonstrated on simulated and real digital angiographic image sequences. It is shown that motion estimation using elastic registration of skeletons is very successful especially with low contrast and noisy angiographic images.

An adaptive regularized recursive displacement estimation algorithm is presented. An estimate of the displacement vector field (DVF) is obtained by minimizing the linearized displaced frame difference (DFD) using nu subsets (submasks) of a set of points that belong to a causal neighborhood (mask) around the working point. Assuming that the displacement vector is constant at all points inside the mask, nu systems of equations are formed based on the corresponding submasks. A set theoretic regularization approach is followed for solving this system of equations by using information about the noise and the solution. An expression for the variance of the linearization error is derived in quantifying the information about the noise. Prior information about the solution is incorporated into the algorithm using a causal oriented smoothness constraint (OSC) which also provides a spatially adaptive prediction model for the estimation DVF. It is shown that certain existing regularized recursive algorithms are special cases of the proposed algorithm, if a single mask is considered. Based on experiments with typical videoconferencing scenes, the improved performance of the proposed algorithm with respect to accuracy, robustness to occlusion and smoothness of the estimated DVF is demonstrated.

An adaptive multiple-input pel-recursive algorithm is presented. The displacement vector field (DVF) is estimated by minimizing the linearized displaced frame difference (DFD) using nu submasks of causal mask around the working point. Then, nu corresponding systems of equations are formed and the set theoretic regularization approach results in a weighted constrained least-squares estimation of the DVF by using information about the variance of the linearization error (noise) and the solution. The prior information about the solution is incorporated into the algorithm using a causal oriented smoothness constraint (OSC), which also provides a spatial prediction model for the estimated DVF. The improved performance of the proposed algorithm with respect to accuracy, robustness to occlusion, and smoothness of the estimated DVF is demonstrated.>

An approach to constrained recursive estimation of the displacement vector field (DVF) in image sequences is presented. An estimate of the displacement vector at the working point is obtained by minimizing the linearized displaced frame difference based on a set of observations that belong to a causal neighborhood (mask). An expression for the variance of the linearization error (noise) is obtained. Because the estimation of the DVF is an ill-posed problem, the solution is constrained by considering an autoregressive (AR) model for the DVF. A nonstationary AR model of the DVF is also considered. Additional information about the solution is incorporated into the algorithm using a causal oriented smoothness constraint. A set theoretic regularization approach based on this formulation results in a weighted constrained least-squares estimation of the DVF. The algorithm shows an improved performance with respect to accuracy, robustness of occlusion, and smoothness of the estimated DVF when applied to typical videoconferencing scenes.>

A model-based approach to pel-recursive motion estimation is presented. The derivation of the algorithm is similar to the Wiener-based pel-recursive motion-estimation algorithm. However, the proposed algorithm utilizes the spatio-temporal correlations in an image sequence by considering an autoregressive (AR) model for the motion-compensated frames. Therefore, depending on the support of the AR model, the estimation is based on two or more consecutive frames. Pel-recursive motion-estimation algorithms which appear in literature are special cases of this algorithm. Various implementation issues such as adaptive regularization and modeling of the motion field are considered. Based on experiments with typical videoconferencing scenes, the conclusion is that the proposed algorithm performs better than the two-frame Wiener-based pel-recursive algorithm with respect to accuracy, robustness, and smoothness of the velocity field.>

In this paper, a distributed detection approach for displacement estimation in image sequences is presented. This method is derived from a Bayesian framework and reduces to a M-ary Hypothesis test among a representative set of possible displacement vectors. It is shown that the mean-squared error based block-matching (BM) algorithm is a special case of this general approach. In our approach, at each point of the current frame a set of overlapping localized detectors outputs a number of estimates for the displacement vector. Then, a distributed detection network is adopted for the fusion of the these estimates. Since the computational load is high, suboptimal but computationally efficient solutions are proposed. The above method gives a more accurate estimation of the displacement field and it is shown to be more robust in the presence of occlusion and noise, compared to the BM algorithm. Experimental results on video-conference image sequences are presented.

A class of iterative image restoration algorithms is derived. The algorithms are based on a representation theorem for the generalized inverse of a matrix. These algorithms exhibit a first or higher order of convergence and some of them consist of an online and an offline computation part. The conditions for convergence and the range of convergence of these algorithms are derived. An iterative algorithm is also presented which exhibits a higher rate of convergence than the standard quadratic algorithm with no extra computational load. These algorithms can be applied to the restoration of signals of any dimensionality. Iterative restoration algorithms that have appeared in the literature represent special cases of the class of algorithms described. Therefore, the approach presented unifies a large number of iterative restoration algorithms.>