Abstract: A method and system for segmentation and removal of pulmonary arteries, pulmonary veins, and a left atrial appendage from 3D medical image data, such as 3D computed tomography (CT) volumes, is disclosed. A global shape model is segmented for each of pulmonary arteries, pulmonary veins, and a left atrial appendage in a 3D volume. The segmented global shape model for each of the pulmonary arteries, pulmonary veins, and left atrial appendage is locally refined based in local voxel intensities in the 3D volume, resulting in a respective mask for each structure. The mask is used to remove voxels belonging to the pulmonary arteries, pulmonary veins, and left atrial appendage from the 3D volume in order to better visualize coronary arteries and bypass arteries.

Abstract: A method and system for adaptive discriminant learning and measurement fusion for image based catheter tracking is disclosed. An adaptive discriminant model is trained online based on a tracked object, such as a pigtail catheter tip, in at least one previous frame of a fluoroscopic image sequence. The object is tracked in the current frame of the fluoroscopic image sequence based at least on the adaptive discriminant model trained online. The object may be tracked in the current frame based on a fusion of three types of measurement models including the adaptive discriminant model trained online, an object detection model trained offline, and an online appearance model.

Abstract: A system and method for performing shape-constrained aortic valve landmark detection using 3D medical images is provided. A rigid global shape defining initial positions of a plurality of aortic valve landmarks is detected within a 3D image. Each of the plurality of aortic valve landmarks is detected based on the initial positions.

Abstract: A method and system for estimating 3D cardiac motion from a single C-arm angiography scan is disclosed. An initial 3D volume is reconstructed from a plurality of 2D projection images acquired in a single C-arm scan. A static mesh is extracted by segmenting an object in the initial 3D volume. The static mesh is projected to each of the 2D projection images. A cardiac phase is determined for each of the 2D projection images. A deformed mesh is generated for each of a plurality of cardiac phases based on a 2D contour of the object and the projected mesh in each of the 2D projection images of that cardiac phase.

Abstract: A computer-implemented method of determining an interatrial septum ring in a cardiac image includes determining a left atrium mean shape based on a plurality of training images and determining an interatrial septum ring mean shape based on the left atrium mean shape. A left atrium mesh is identified in a new image. Then, a deformation field from the left atrium mean shape to the left atrium mesh is calculated and applied to the interatrial septum ring mean shape to determine the interatrial septum ring in the new image.

Abstract: A method is disclosed for segmentation of a calcified blood vessel in image data. An embodiment of the method includes providing a vesseltree representation of the blood vessel; providing a number of preliminary boundary representations of a number of cross-sections of the blood vessel; providing a number of intensity profiles in the image data in the number of cross-sections; determining a calcification in the cross-section based on the intensity profile; and correcting each preliminary boundary representation into a corrected boundary representation which excludes the calcification from an inner part of the blood vessel. A segmentation system is also disclosed.

Abstract: A method and a segmentation system are disclosed. An embodiment of the method includes providing an image representation of the structure; providing a start surface model, including a mesh with a plurality of vertices connected by edges; defining for each vertex a ray normal to the surface model at the position of the vertex; assigning more than two labels to each vertex, each label representing a candidate position of the vertex on the ray; providing a representation of likelihoods for each candidate position the likelihood referring to whether the candidate position corresponds to a surface point of the structure in the image representation; and defining a first order Markow Random Field with discrete multivariate random variables, the random variables including the labels of the candidate positions and the representation of likelihoods, finding an optimal segmentation of the structure by using an maximum a posteriori estimation in this Markow Random Field.

Abstract: A method of performing image retrieval includes training a random forest RF classifier based on low-level features of training images and a high-level feature, using similarity values generated by the RF classifier to determine a subset of the training images that are most similar to one another, and classifying input images for the high-level feature using the RF classifier and the determined subset of images.

Abstract: A method and system for multi-part left atrium (LA) segmentation in a C-arm CT volume is disclosed. Multiple LA part models, including an LA chamber body mesh, an appendage mesh, a left inferior pulmonary vein (PV) mesh, a left superior PV mesh, a right inferior PV mesh, and a right superior PV mesh, are segmented in a 3D volume. A volume mask is generated from the LA chamber mesh, the appendage mesh, and the PV meshes. Erosion is performed in the LA chamber body and a plurality of ostia regions in the volume mask. The plurality of ostia regions in the volume mask are refined using region growing, and a smooth mesh is fit to each ostia region. A consolidated LA mesh is generated from the volume mask and the parts of the LA mesh are relabeled in the ostia region based on part boundaries detected using an optimization approach.

Abstract: A method for displaying a motion compensated overlay includes receiving a model of a structure of interest, capturing an image depicting a region of interest and an instrument, determining whether the structure of interest is visible in the image, registering the model of the structure of interest to the image upon determining that the structure of interest is visible, and combining the model of the structure of interest with the image according to a registration to determine an overlay image.

Abstract: A method and system for patient-specific modeling of the whole heart anatomy, dynamics, hemodynamics, and fluid structure interaction from 4D medical image data is disclosed. The anatomy and dynamics of the heart are determined by estimating patient-specific parameters of a physiological model of the heart from the 4D medical image data for a patient. The patient-specific anatomy and dynamics are used as input to a 3D Navier-Stokes solver that derives realistic hemodynamics, constrained by the local anatomy, along the entire heart cycle. Fluid structure interactions are determined iteratively over the heart cycle by simulating the blood flow at a given time step and calculating the deformation of the heart structure based on the simulated blood flow, such that the deformation of the heart structure is used in the simulation of the blood flow at the next time step.

Abstract: A method and system for isolating the heart in a 3D volume, such as a cardiac CT volume, for patients with coronary artery bypasses is disclosed. An initial heart isolation mask is extracted from a 3D volume, such as a cardiac CT volume. The aortic root and ascending aorta are segmented in the 3D volume, resulting in an aorta mesh. The aorta mesh is expanded to include bypass coronary arteries. An expanded heart isolation mask is generated by combining the initial heart isolation mask with an expanded aorta mask defined by the expanded aorta mesh.

Abstract: A method and system for multi-part left atrium (LA) segmentation in a C-arm CT volume is disclosed. Multiple LA part models, including an LA chamber body mesh, an appendage mesh, a left inferior pulmonary vein (PV) mesh, a left superior PV mesh, a right inferior PV mesh, and a right superior PV mesh, are segmented in a 3D volume. The LA chamber body mesh and the appendage mesh may be segmented as a combined object and the PV meshes may be segmented subject to a statistical shape constraint. A consolidated LA mesh is generated from the segmented LA part models.

Abstract: A method and system for coronary artery detection in 3D cardiac volumes is disclosed. The heart chambers are segmented in the cardiac volume, and an initial estimation of a coronary artery is generated based on the segmented heart chambers. The initial estimation of the coronary artery is then refined based on local information in the cardiac volume in order to detect the coronary artery in the cardiac volume. The detected coronary artery can be extended using 3D dynamic programming.

Abstract: A method and system for model based fusion pre-operative image data, such as computed tomography (CT), and intra-operative C-arm CT is disclosed. A first pericardium model is segmented in the pre-operative image data and a second pericardium model is segmented in a C-arm CT volume. A deformation field is estimated between the first pericardium model and the second pericardium model. A model of a target cardiac structure, such as a heart chamber model or an aorta model, extracted from the pre-operative image data is fused with the C-arm CT volume based on the estimated deformation field between the first pericardium model and the second pericardium model. An intelligent weighted average may be used improve the model based fusion results using models of the target cardiac structure extracted from pre-operative image data of patients other than a current patient.

Abstract: A method for displaying a motion compensated overlay includes receiving a model of a structure of interest, capturing an image depicting a region of interest and an instrument, determining whether the structure of interest is visible in the image, registering the model of the structure of interest to the image upon determining that the structure of interest is visible, and combining the model of the structure of interest with the image according to a registration to determine an overlay image.

Abstract: A method and system for intraoperative guidance in an off-pump mitral valve repair procedure is disclosed. A plurality of patient-specific models of the mitral valve are generated, each from pre-operative image data obtained using a separate imaging modality. The pre-operative image data from the separate imaging modalities are fused into a common coordinate system by registering the plurality of patient-specific models. A model of the mitral valve is estimated in real-time in intraoperative image data using a fused physiological prior resulting from the registering of the plurality of patient-specific models.

Abstract: A method and system for providing detecting and classifying coronary stenoses in 3D CT image data is disclosed. Centerlines of coronary vessels are extracted from the CT image data. Non-vessel regions are detected and removed from the coronary vessel centerlines. The cross-section area of the lumen is estimated based on the coronary vessel centerlines using a trained regression function. Stenosis candidates are detected in the coronary vessels based on the estimated lumen cross-section area, and the significant stenosis candidates are automatically classified as calcified, non-calcified, or mixed.

Abstract: A method for guiding transcatheter aortic valve implantations includes receiving an interventional 3D image of an aortic root reconstructed from a sequence of 2D images acquired from a C-arm computed tomography (CT) system being rotated about a patient through a predetermined number of degrees, segmenting the aortic root and detecting aortic root landmarks in the 3D image, where the aortic root landmarks include three lowest points of aortic root cusps, two coronary artery ostia, and three commissures points where the cusps meet, cropping an area inside the segmented aortic root out of the 3D volume for volume rendering, centering the 3D image on an intersection of two orthogonal planes, each containing the two detected coronary ostia, that are orthogonal to a plane spanned by three lowest points of the aortic root cusps, and volume rendering the 3D cropped aortic root image together with the detected landmarks onto a 2D image.

Abstract: A method and system for automatic extraction of personalized left atrium models is disclosed. A left atrium chamber body is segmented from a 3D image volume. At least one pulmonary venous ostium is detected on the segmented left atrium chamber body. At least one pulmonary vein trunk connected to the left atrium chamber body is segmented based on the detected pulmonary venous ostia.