Abstract: The present invention relates to a device for segmenting an image (12) of a subject (14) comprising a data interface (16) for receiving an image (12) of said subject (14) and at least one contour (18) or at least one part of a contour (18), said contour (18) indicating a structure (19) within said image (12), a selection unit (20) for selecting a region (22) in said image (12) divided into a first and a second disjoint part (24, 26) by said contour (18) or said part of said contour (18), said selected region (22) comprising a drawn region and/or a computed region, a classifier (28) for classifying said selected region (22) based on at least one parameter for image segmentation, an analysis unit (29) for defining an objective function based on the classification result, an optimizer (30) for optimizing said parameter set by varying an output of said objective function and an image segmentation unit (32) for segmenting said image (12) using said optimized parameter set.

Abstract: A method includes obtaining a single training image from a set of training images in a data repository. The method further includes generating an initial tissue class atlas based on the obtained single training image. The initial tissue class atlas includes two or more different tissue class images corresponding to two or more different tissue classes. The method further includes registering the remaining training images of the set of training images to the initial tissue class atlas. The method further includes generating a quality metric for each of the registered images. The method further includes evaluating the quality metric of each of the registered image with a predetermined evaluation criterion. The method further includes identifying a sub-set of images from the set of training images that satisfy the evaluation criterion. The method further includes generating a subsequent tissue class atlas based on the identified sub-set of the set of training images.

Abstract: The present invention relates to a device for segmenting an image (12) of a subject (14) comprising a data interface (16) for receiving an image (12) of said subject (14) and at least one contour (18) or at least one part of a contour (18), said contour (18) indicating a structure (19) within said image (12), a selection unit (20) for selecting a region(22) in said image (12) divided into a first and a second disjoint part (24, 26) by said contour (18) or said part of said contour (18),said selected region (22) comprising a drawn region and/or a computed region, a classifier (28) for classifying said selected region (22) based on at least one parameter for image segmentation, an analysis unit (29) for defining an objective function based on the classification result, an optimizer (30) for optimizing said parameter set by varying an output of said objective function and an image segmentation unit (32) for segmenting said image (12) using said optimized parameter set.

Abstract: A radiation planning system includes a predictor-corrector optimizer unit which computes a predicted dose based on a collection of control points with a current approximate dose, each control point with a corresponding set of leaf positions, and determines an additional control point with a corresponding set of leaf positions based on a difference of the predicted fluence and the current approximate fluence through a least cost or shortest path in a layered graph structure of realizable leaf positions. Tools are described to help a planner to evaluate the effect of parameter changes to the current plan based on an identified zone of influence. The planner interactively views the current plan based on a visualization of the plan objectives and correlations between the objectives.

Abstract: Interactive mesh deformation for in-plane 3D segmentation/delineation for radiation therapy planning done on a slice by slice basis of a region/a volume of interest (VOI, ROI). Segmentation starts by some automatic 3D algorithm approximating the organ surface roughly by some triangular surface mesh which mesh is afterwards manually refined by the user who deforms it to bring it closer to the region of interest. The deformation is an invertible, i.e. one-to-one, mapping avoiding self-intersections of the deformed mesh thereby preserving the topology of the anatomy. The deformation mapping involves a Gaussian function (Gaussian deformation kernel) restricting the deformation to a local region. The user picks with the pointer a start point on a selected image slice through the volume and moves it to some end point. The distance the mesh vertices move decreases exponentially with the distance to the start point.

Abstract: The invention relates to a labeling apparatus (1) for labeling structures of an object shown in an object image. A probability map providing unit (3) provides a probability map, the probability map indicating for different labels, which are indicative of different structures of the object, and for different positions in the probability map the probability that the respective structure, which is indicated by the respective label, is present at the respective position, wherein the probability depends on the position in the probability map. The probability map is mapped to the object image by a mapping unit (4), wherein a label assigning unit (5) assigns to a provided contour, which represents a structure in the object image, a label based on the mapped probability map. This allows automatically labeling structures of the object, which are indicated by provided contours in the object image, with relatively low computational efforts.

Abstract: The present application is directed to the idea of using sampling techniques to propagate segmentation uncertainty in order to evaluate variability in radiation planning. A radiotherapy planning apparatus (10) creates diagnostic image data of a region of interest of a subject. Image data from other sources can also be used. The image data is segmented (44) and combined with previously imaged model data. Target measures such as dose volume histograms are produced for each of the segmentations of the image data. These measures are later combined into a statistical quantification of the target measure (FIG. 3). This information is presented (52) to the user to give the user possible outcomes of the radiotherapy plan, and, e.g., confidence levels in those outcomes.

Abstract: A method includes obtaining a single training image from a set of training images in a data repository. The method further includes generating an initial tissue class atlas based on the obtained single training image. The initial tissue class atlas includes two or more different tissue class images corresponding to two or more different tissue classes. The method further includes registering the remaining training images of the set of training images to the initial tissue class atlas. The method further includes generating a quality metric for each of the registered images. The method further includes evaluating the quality metric of each of the registered image with a predetermined evaluation criterion. The method further includes identifying a sub-set of images from the set of training images that satisfy the evaluation criterion. The method further includes generating a subsequent tissue class atlas based on the identified sub-set of the set of training images.

Abstract: A magnetic resonance (MR) system (10) and method (100) maintains geometric alignment of diagnostic scans during an examination of a patient (12). At least one processor (40) is programmed to, in response to repositioning of the patient (12) during the examination, perform an updated survey scan of the patient (12). A scan completed during the examination is selected as a template scan. A transformation map between the template scan and the updated survey scan is determined using a registration algorithm, and the transformation map is applied to a scan geometry of a remaining diagnostic scan of the examination. A scan plan for the remaining diagnostic scan is generated using the updated scan geometry. The remaining diagnostic scan is performed according to the scan plan.

Abstract: The invention relates to a labeling apparatus (1) for labeling structures of an object shown in an object image. A probability map providing unit (3) provides a probability map, the probability map indicating for different labels, which are indicative of different structures of the object, and for different positions in the probability map the probability that the respective structure, which is indicated by the respective label, is present at the respective position, wherein the probability depends on the position in the probability map. The probability map is mapped to the object image by a mapping unit (4), wherein a label assigning unit (5) assigns to a provided contour, which represents a structure in the object image, a label based on the mapped probability map. This allows automatically labeling structures of the object, which are indicated by provided contours in the object image, with relatively low computational efforts.

Abstract: When modeling anatomical structures in a patient for diagnosis or therapeutic planning, an atlas (26) of segmented co-registered CT and MRI anatomical structure reference images can be accessed, and an image of one or more such structures can be selected and overlaid with an MRI image of corresponding structure(s) in a clinical image of a patient. A user can click and drag landmarks or segment edges on the reference MRI image to deform the reference MRI image to align with the patient MRI image. Registration of a landmark in the patient MRI image to the reference MRI image also registers the patient MRI image landmark with a corresponding landmark in the co-registered reference CT image, and electron density information from the reference CT image landmark is automatically attributed to the corresponding registered patient MRI landmark.

Abstract: When adapting models of anatomical structures in a patient for diagnosis or therapeutic planning, an atlas (26) of predesigned anatomical structure models or image volumes can be accessed, and a segmentation of one or more such structures can be selected and overlaid on an a 3D image of corresponding structure(s) in a clinical image (52) of a patient. A user can click on an initially unapproved segmentation 5 landmark (72) on the patient image (52), reposition the unapproved landmark, and approve the repositioned landmark. Remaining unapproved landmarks (72) are then repositioned as a function of the position of the approved landmark (92) using one or more interpolation techniques to adapt the model to the patient image on the fly.

Abstract: A method includes generating a set of group-wise registered images from a time sequence of images based on a region of interest of a subject or object identified in at least one of the images, the image sequence, and a motion model indicative of an estimate of a motion of the subject or object during which the image sequence is acquired.

Abstract: When modeling anatomical structures in a patient for diagnosis or therapeutic planning, an atlas (26) of predesigned anatomical structure models can be accessed, and model of one or more such structures can be selected and overlaid on an a 3D image of corresponding structure(s) in a clinic image of a patient. A user can click and drag a cursor on the model to deform the model to align with the clinical image. Additionally, a processor (16) can generate a volumetric deformation function using splines, parametric techniques, or the like, and can deform the model to fit the image in real time, in response to user manipulation of the model.

Abstract: Interactive mesh deformation for in-plane 3D segmentation/delineation for radiation therapy planning done on a slice by slice basis of a region/a volume of interest (VOI, ROI). Segmentation starts by some automatic 3D algorithm approximating the organ surface roughly by some triangular surface mesh which mesh is afterwards manually refined by the user who deforms it to bring it closer to the region of interest. The deformation is an invertible, i.e. one-to-one, mapping avoiding self-intersections of the deformed mesh thereby preserving the topology of the anatomy. The deformation mapping involves a Gaussian function (Gaussian deformation kernel) restricting the deformation to a local region. The user picks with the pointer a start point on a selected image slice through the volume and moves it to some end point. The distance the mesh vertices move decreases exponentially with the distance to the start point.

Abstract: A method for segmenting image data includes identifying a 2D boundary start position corresponding to tissue of interest in a cross-section of volumetric image data, wherein the start position is identified by a current position of a graphical pointer with respect to the cross-section, generating a preview 2D boundary for the tissue of interest based on the start position, displaying the preview 2D boundary superimposed over the cross-section, and updating the displayed preview 2D boundary if the position of the graphical pointer changes with respect to the cross-section.

Abstract: A system and method for segmenting an anatomical structure. The system and method initiating a segmentation algorithm, which produces a surface mesh of the anatomical structure from a series of volumetric images, the surface mesh formed of a plurality of polygons including vertices and edges, assigning a spring to each of the edges and a mass point to each of the vertices of the surface mesh, displaying a 2D reformatted view including a 2D view of the surface mesh and the anatomical structure, adding pull springs to the surface mesh, the pull springs added based upon a selected point on a surface of the surface mesh and moving a portion of the surface mesh via an interactive point.

Abstract: A system and method for segmenting an image of an organ. The system and method including selecting a surface model of the organ, selecting a plurality of points on a surface of an image of the organ and transforming the surface model to the plurality of points on the image.

Abstract: A method includes generating a set of group-wise registered images from a time sequence of images based on a region of interest of a subject or object identified in at least one of the images, the image sequence, and a motion model indicative of an estimate of a motion of the subject or object during which the image sequence is acquired.

Abstract: When modeling anatomical structures in a patient for diagnosis or therapeutic planning, an atlas (26) of segmented co-registered CT and MRI anatomical structure reference images can be accessed, and an image of one or more such structures can be selected and overlaid with an MRI image of corresponding structure(s) in a clinical image of a patient. A user can click and drag landmarks or segment edges on the reference MRI image to deform the reference MRI image to align with the patient MRI image. Registration of a landmark in the patient MRI image to the reference MRI image also registers the patient MRI image landmark with a corresponding landmark in the co-registered reference CT image, and electron density information from the reference CT image landmark is automatically attributed to the corresponding registered patient MRI landmark.