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 assigning first localization data of a breast of a patient derived from first image data of the breast, the first image data being the result of a first radiological data acquisition process, to second localization data of the same breast derived from second image data, the second image data being the result of a second radiological data acquisition process, or vice versa. Thereby, the first localization data are assigned to the second localization data by intermediately mapping them into breast model data representing a patient-specific breast shape of the patient and then onto the second image data—or vice versa, thereby deriving assignment data. An assignment system performs the above-described method.

Abstract: A method in radiological imaging for determining lesions in image data of an examination object is described. In an embodiment, the method includes determining anatomical structures by hierarchical breakdown of the image data of the examination object. The method furthermore includes image data analysis for localizing lesion candidates in the anatomical structures. Moreover, the method also includes determining the lesions by evaluating and filtering the lesion candidates. Moreover, an image processing workstation in radiological imaging for determining lesions in image data of an examination object and an imaging apparatus are described.

Abstract: A method and system for on-line learning of landmark detection models for end-user specific diagnostic image reading is disclosed. A selection of a landmark to be detected in a 3D medical image is received. A current landmark detection result for the selected landmark in the 3D medical image is determined by automatically detecting the selected landmark in the 3D medical image using a stored landmark detection model corresponding to the selected landmark or by receiving a manual annotation of the selected landmark in the 3D medical image. The stored landmark detection model corresponding to the selected landmark is then updated based on the current landmark detection result for the selected landmark in the 3D medical image. The landmark selected in the 3D medical image can be a set of landmarks defining a custom view of the 3D medical image.

Abstract: An embodiment of the method is disclosed for non-invasive lesion candidate detection in a patient's body includes generating a number of first medical images of the patient's body. The method further includes identifying lesion-like geometrical regions inside the first medical images of the patient's body by applying image processing methods, whereby the identification is at least partly controlled by a number of patient-specific context features which are not directly extractable from the first medical images. In addition, the method includes selecting a number of the identified lesion-like geometrical regions as lesion candidates.

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 automatic detection and volumetric quantification of bone lesions in 3D medical images, such as 3D computed tomography (CT) volumes, is disclosed. Regions of interest corresponding to bone regions are detected in a 3D medical image. Bone lesions are detected in the regions of interest using a cascade of trained detectors. The cascade of trained detectors automatically detects lesion centers and then estimates lesion size in all three spatial axes. A hierarchical multi-scale approach is used to detect bone lesions using a cascade of detectors on multiple levels of a resolution pyramid of the 3D medical image.

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 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 automatic detection and volumetric quantification of bone lesions in 3D medical images, such as 3D computed tomography (CT) volumes, is disclosed. Regions of interest corresponding to bone regions are detected in a 3D medical image. Bone lesions are detected in the regions of interest using a cascade of trained detectors. The cascade of trained detectors automatically detects lesion centers and then estimates lesion size in all three spatial axes. A hierarchical multi-scale approach is used to detect bone lesions using a cascade of detectors on multiple levels of a resolution pyramid of the 3D medical image.

Abstract: A method and system for the diagnosis of 3D images are disclosed, which significantly cuts the time required for the diagnosis. The 3D images are for example an image volume dataset of a magnetic resonance tomography system which is saved in an RIS or PACS system. In at least one embodiment, the diagnostic finding are partially automatically generated, and details of the position, size and change in pathological structures are compared to previous diagnostic findings are generated automatically. As a result of this automation the diagnostic work of radiologists is significantly reduced.

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 and a system are disclosed for image annotation of images, in particular two- and three-dimensional medical images. In at least one embodiment, the image annotation system includes an image parser which parses images retrieved from an image database or provided by an image acquisition apparatus and segments each image into image regions. The image can be provided by any kind of image acquisition apparatus such as a digital camera an x-ray apparatus, a computer tomograph or a magnetic resonance scanning apparatus. Each segmented image regions is annotated automatically with annotation data and stored in an annotation database. In at least one embodiment, the system includes at least one user terminal which loads at least one selected image from said image database and retrieved the corresponding annotation data of all segmented image regions of said image from said annotation database for further annotation of the image.

Abstract: A method for training a computer system for automatic detection of regions of interest includes receiving patient records. For each of the received patient records a text field and a medical image are identified from within the patient record and the medical image is automatically segmented to identify a structure of interest. The text field is searched for one or more keywords indicative of a particular abnormality associated with the structure of interest. The medical image is added to a grouping representing the particular abnormality when the text field indicates that the patient has the particular abnormality and the medical image is added to a grouping representing the absence of the particular abnormality when the text field does not indicate that the patient has the particular abnormality. The groupings of medical images are used to automatically train a computer system for the subsequent detection of the particular abnormality.

Abstract: A method and system for fully automatic segmentation the prostate in multi-spectral 3D magnetic resonance (MR) image data having one or more scalar intensity values per voxel is disclosed. After intensity standardization of multi-spectral 3D MR image data, a prostate boundary is detected in the multi-spectral 3D MR image data using marginal space learning (MSL). The detected prostate boundary is refined using one or more trained boundary detectors. The detected prostate boundary can be split into patches corresponding to anatomical regions of the prostate and the detected prostate boundary can be refined using trained boundary detectors corresponding to the patches.

Abstract: A method and apparatus for automatic detection and labeling of 3D spinal geometry is disclosed. Cervical, thoracic, and lumbar spine regions are detected in a 3D image. Intervertebral disk candidates are detected in each of the spine regions using iterative marginal space learning (MSL). Using a global probabilistic spine model, a separate one of the intervertebral disk candidates is selected for each of a plurality of labeled intervertebral disk locations.

Abstract: A method for training a computer system for automatic detection of regions of interest includes receiving patient records. For each of the received patient records a text field and a medical image are identified from within the patient record and the medical image is automatically segmented to identify a structure of interest. The text field is searched for one or more keywords indicative of a particular abnormality associated with the structure of interest. The medical image is added to a grouping representing the particular abnormality when the text field indicates that the patient has the particular abnormality and the medical image is added to a grouping representing the absence of the particular abnormality when the text field does not indicate that the patient has the particular abnormality. The groupings of medical images are used to automatically train a computer system for the subsequent detection of the particular abnormality.