Fast detection and characterization of vessels in very large 3-D data sets using geometrical moments.

An improved and very fast algorithm dealing with the extraction of vessels in three-dimensional imaging is described. The approach is based on geometrical moments and a local cylindrical approximation. A robust estimation of vessel and background intensity levels, position, orientation, and diameter of the vessels with adaptive control of key parameters, is provided during vessel tracking. Experimental results are presented for lower limb arteries in multidetector computed tomography scanner.

Future trends in 3D medical imaging.

Areas where significant progress is needed are outlined. A unified conceptual framework based on generic tasks, showing the intricacies and the dependences that exist among completely distinct and intensive research lines, is defined. True 3-D volume imaging devices are then discussed, with the main emphasis on vascular network reconstruction. Segmentation issues are briefly reviewed. The fusion of information dealing with image, signal, and model is described. Simulation and planning problems are discussed.

Spatio-temporal grouping for the formation of vascular segments in coronarography image sequences.

This work is aimed at the extraction of vascular structures using perceptual grouping methods. The approach is based on structural feature matching, between images in the time sequence. These features correspond to junctions (or bifurcation points) and are paired by means of a local criteria. A global refinement procedure is then applied by means of a relaxation scheme. The feature trajectories over the entire image sequence are examined to further build sets of vascular segments, which better describe vascular branches. The resulting performances are exemplified on a standard biplane coronarographic examination.

Generating anatomical models of the heart and the aorta from medical images for personalized physiological simulations.

The anatomy and motion of the heart and the aorta are essential for patient-specific simulations of cardiac electrophysiology, wall mechanics and hemodynamics. Within the European integrated project euHeart, algorithms have been developed that allow to efficiently generate patient-specific anatomical models from medical images from multiple imaging modalities. These models, for instance, account for myocardial deformation, cardiac wall motion, and patient-specific tissue information like myocardial scar location. Furthermore, integration of algorithms for anatomy extraction and physiological simulations has been brought forward. Physiological simulations are linked closer to anatomical models by encoding tissue properties, like the muscle fibers, into segmentation meshes. Biophysical constraints are also utilized in combination with image analysis to assess tissue properties. Both examples show directions of how physiological simulations could provide new challenges and stimuli for image analysis research in the future.

An improved spatial tracking algorithm applied to coronary veins into Cardiac Multi-Slice Computed Tomography volume.

This paper describes an enhanced vessel tracking algorithm. The method specifity relies on the coronary venous tree extraction through Cardiac Multi-Slice Computed Tomography (MSCT). Indeed, contrast inhomogeneities are a major issue in the data sets that necessitate a robust tracking procedure. The method is based on an existing moment-based algorithm designed for coronary arteries into MSCT volume. In order to extract the whole path of interest, improvements concerning progression strategy are proposed. Furthermore, the original procedure is combined with an automatic recentring method based on ray casting. This enhanced method has been tested on three data sets. According to the first results, the method appears robust to curvatures, contrast inhomogeneities and low contrast blood veins.

Temporal tracking of coronaries in multi-slice computed tomography.

A method is proposed that performs a temporal tracking of the coronaries in multi-slice computed tomography (MSCT) dynamic sequences. The process exploits geometric moments and a local cylindrical approximation of the vessel to estimate the local characteristics of the structure in each volume and estimate its displacement along the sequence. The research strategy is based on a region matching process to find the location of the point in the successive volumes. A spatial tracking is then applied to refine its location inside the vessel. Tests have been achieved on simulated and real displacements of coronary segments.

Simulation environment for the evaluation of 3D coronary tree reconstruction algorithms in rotational angiography.

We present a preliminary version of a simulation environment to evaluate the 3D reconstruction algorithms of the coronary arteries in rotational angiography. It includes the construction of a 3D dynamic model of the coronary tree from patient data, the modeling of the rotational angiography acquisition system to simulate different acquisition and gating strategies and the calculation of radiographic projections of the 3D model of coronary tree throughout several cardiac cycles.

Temporal tracking of coronaries in MSCTA by means of 3D geometrical moments.

An algorithm is proposed that perform a temporal tracking of the vessel central axis in a 3-D dynamic sequence in multi-slice computed tomography (MSCT). The approach is based on geometric moments and a local cylindrical approximation. The local characteristics of the vessel are estimated on the first volume of the sequence (position on the vessel central axis, local diameter, intravascular and background intensities), then used to track the vessel along the sequence. The correspondence between two volumes is solved through a region matching based on a criterion of minimal distance combining moment-based descriptors with intensity information. Preliminary results are presented on two sequences.

A multiscale tracking algorithm for the coronary extraction in MSCT angiography.

This paper deals with the extraction of the coronary network on dynamic volume sequences, acquired in multi-slice spiral computed tomography (MSCT). The proposed approach makes use of a tracking algorithm of the vascular structure, combining a 3D geometric moment operator with a multiscale Hessian filter to estimate the vessel central axis location, its local diameter and orientation. The method performs at the same time, a bifurcation detection to reconstitute the structure of the coronary network. The mean computation time to extract a coronary network is about 3 minutes using a P4-2.4G PC. Preliminary encouraging results are presented on one volume of a sequence.

A 3D Static Heart Model From a MSCT Data Set.

Dynamic Computed Tomography (CT) imaging aims to access the kinetics of the moving organs. In cardiac imaging, the interest lies in the possibility of obtaining anatomic and functional information on the heart and the coronaries during the same examination. However, segmentation, reconstruction and registration algorithms need to be developed for diagnostic purposes. We propose thus to built a 3D heart model from Multi-slice Spiral Computed Tomography (MSCT) dynamic sequences to facilitate the evaluation of these algorithms. The model building relies on semi-automatic segmentation techniques based on deformable models such as Fast Marching and active contours. Shape-based interpolation and Marching Cube algorithms are then used for the 3D surface reconstruction.

Vessel extraction in coronary X-ray Angiography.

This paper describes a method to extract the vascular centerlines and contours in coronary angiography. The proposed approach associates geometric moments for the estimation of a "cylinder-like model" and relies on a tracking process. The orientation of the cylinder axis and its local diameter are computed from the analytical expressions of the geometric moments of up to order 2. Experimental results are presented on several images of two sequences that show the efficiency of the method.

Coronary extraction and stenosis quantification in X-ray angiographic imaging.

This work describes a method to quantify stenosis in X-ray coronary angiography. Vascular edge extraction is first performed based on a deformable spline algorithm. It makes use of directional S-Gabor filters to build an external energy field that is then used in a snake optimisation scheme. A string matching technique is then applied to match the contour points and obtain a trace between the matched points. This trace allows then computing the vessel diameter and deriving quantitative stenosis measurements. Experimental results are presented on simulated data and real images.

Vascular network segmentation in subtraction angiograms: a comparative study.

The three-dimensional reconstruction of vascular trees from a very limited number of two-dimensional projections is an active research field. The quality of the results (resolution in terms of vessel size and geometrical location) is highly dependent on the overall distortions introduced in the data acquisition process but is also related to the reliability of feature detection. A new approach based on mathematical morphology is proposed in this paper for extracting the centrelines and edges of coronary vessels from digital subtracted angiograms. These results are compared with those from an alternative method based on vectorial tracking and a directed contour finder.