Siemens Aktiengesellschaft

Siemens Medical Solutions USA, Inc.

Yiyong Sun

Chenyang Xu

Frank Sauer

Rui Liao

Xiaolei Sharon Huang

Norbert Rahn

Liron Yatziv

Herve Jun Alfred Lombaert

Herve Lombaert

Leo John Grady

Scott Bradley Baker

Matthias John

Christophe Chefd'hotel

Ali Khamene

Jens Gühring

Christoph Guetter

Sebastian Vogt

Fred S Azar

Farida Cheriet

Luc Duong

Dieter Hahn

A multilevel image segmentation technique using graph cuts is disclosed. A reduced resolution image is generated from a full resolution image which is to be segmented. The reduced resolution image is then segmented in order to identify a boundary between an object and a background within the image. The identified boundary then identifies a portion of an increased resolution image which is segmented in order to refine the earlier identified boundary. The steps may be iterated for successively increasing image resolutions in order to refine the boundary as required by a particular application. An initial identification of object and background portions of the image may be provided as input by a user. Alternatively, a user may identify only the object portion, and the background portion may be automatically determined.

Multilevel image segmentation

A method for three-dimensional reconstruction of a branched object from a rotational sequence of images of the branched object includes segmenting the branched object from each image of the sequence, extracting centerlines of the branched object, performing symbolic reconstruction via a stereo correspondence matching between the centerlines from different views of the sequence of images using a graph cut-based optimization, and creating a three-dimensional tomographic reconstruction of the branched object compensated for motion of the branched object between the images of the sequence.

Coronary reconstruction from rotational X-ray projection sequence

A method for segmenting at least a pair of regions of an image. High resolution data is obtained of the image. Each one of the pair of the regions in the image is marked. Graph cuts are used on the downsampled data to obtain first voxels along an outer boundary of a selected one of the pair of marked regions and second voxels along an inner boundary the selected region. The graphs cuts are projected to the previously obtained high-resolution image data. First and second sets of seeds are placed on the first voxels and a second set of seeds respectively. The first seeds grow into first areas extending inwardly of the selected region while simultaneously the second seeds grow into second areas extending towards the first extending areas until the first areas and the second areas meet to thereby establish the outer boundary of the selected region.

Fast 4D segmentation of large datasets using graph cuts

A fluoroscopy image is registered with data representing a volume, identifying a catheter position relative to a volume represented by preoperative volume data. The catheter position relative to a patient volume represented by data acquired without scanning the catheter is displayed. For example, a 2D fluoroscopy image is registered with respect to coordinates of a 3D preoperative CT or MRI volume by registering the fluoroscopy image and the preoperative volume to 3D ultrasound coordinates.

Registering 2D and 3D data using 3D ultrasound data

A method for registering images of multiple modalities includes acquiring first image of a subject using a first modality. A second image of the subject is acquired using a second modality. The first image includes greater structural detail of the subject than the second image and the second image is a video image including multiple image frames. The first and second images are registered based on an anatomical structure observable in the first image and a foreign object proximate to the anatomical structure observable in the second image.

Registration of CT volumes with fluoroscopic images

A method of performing 4-dimensional virtual endoscopy includes acquiring a sequence of cardiac images of a patient's heart from an imaging device; acquiring an electrocardiographic signal exhibiting cardiac cycles of the heart; and utilizing the electrocardiographic signal to gate the images for deriving respective 3-dimensional views in succession at corresponding cardiac cycles, from respective cardiac images at a selected phase point common to each of the corresponding cardiac cycles.

Method and apparatus for interactive 4-dimensional (4D) virtual endoscopy

A method of registering 3-dimensional digitized images to 2-dimensional digitized images during a medical procedure includes providing a pair of correctly-registered training images L={l, l} and their joint intensity distribution p(i, i), wherein iand iare reference and floating images, respectively, providing a pair of observed images O={o, o} and their joint intensity distribution p(i, i), mapping a marginal intensity distribution of the observed pair O={o, o} to a marginal intensity distribution of the training pair L={l, l}, and estimating a set of parameters T that registers image oto image oby maximizing a weighted sum of a Jensen-Shannon divergence (JSD) of a joint intensity distribution of the observed pair and a joint intensity distribution of the training pair and a similarity measure between the observed images.

System and method for learning-based 2D/3D rigid registration for image-guided surgery using Jensen-Shannon divergence

An intuitive user interface is provided for endoscopic view visualization. The user interface may permit individual or simultaneous navigation through images displayed in an endoscopic view window, a three dimensional overview window, and a multi-planar reformatted window. An operation performed directly on one of the windows may commence the navigation. The endoscopic view window may be divided into a plurality of sub-regions, each sub-region having a corresponding navigation function. The endoscopic view window and the three dimensional overview window may have corresponding icons related to a single location in corresponding images being displayed. The corresponding icons may be color coded to facilitate intuitive alignment of the corresponding images displayed in the different windows. The three dimensional overview window may have a geometric icon to facilitate intuitive alignment of the corresponding images. The multi-planar reformatted window may permit spherical navigation through the images displayed.

Intuitive user interface for endoscopic view visualization

A computer-implemented method is disclosed for comparing three dimensional (3D) digital medical images. The method uses a reference MPR to position subsequent MPRs in one or more other 3D digital medical images so their content matches the reference MPR. The matched MPRs may then be used by a medical professional to diagnose a patient condition. The ability to quickly and automatically position matching MPRs for multiple 3D images eases the medical staff workload and shortens diagnostics time. Matching MPRs provides an effective way to view the 3D volumes for anatomical changes over time and to monitor medical conditions such as stenosis and tumors.

Automatic positioning of matching multi-planar image reformatting (MPR) views of multiple 3D medical images

A system and method for automatically registering a three dimensional (3D) pre-operative image of an anatomical structure with intra-operative electrophysiological (EP) points of a 3D electro-anatomical (EA) image map of the anatomical structure is disclosed. The pre-operative image is displayed in a first supporting view. The intra-operative EA image map is displayed in a second supporting view. An alignment of the pre-operative image with the intra-operative map is performed by identifying at least one corresponding point on each image. The view of the pre-operative image is integrated with the EA map based on the alignment.

System and method for automatically registering three dimensional cardiac images with electro-anatomical cardiac mapping data

A method and system for non-rigidly registering a fixed to a moving image utilizing a B-Spline based free form deformation (FFD) model is disclosed. The methodology utilizes sparse feature correspondences to estimate an elastic deformation field in a closed form. In a multi-resolution manner, the method is able to recover small to large non-rigid deformations. The resulting deformation field is globally smooth and guarantees one-to-one mapping between the images being registered. The method generally comprises the steps of: detecting feature points on the fixed image and feature points on the moving image; assigning a feature vector to each feature point; calculating the dissimilarity of each pair of feature vectors for feature pairs on the fixed image and the moving image; calculating the correspondence between feature pairs based on the dissimilarity measure; solving for a dense deformation field P using a closed form FFD model; and transforming the moving image and the feature points on the moving image using a current FFD deformation field estimate.

Fast parametric non-rigid image registration based on feature correspondences

A method for aligning a pair of images includes providing a pair of images, identifying salient feature regions in both a first image and a second image, wherein each region is associated with a spatial scale, representing feature regions by a center point of each region, registering the feature points of one image with the feature points of the other image based on local intensities, ordering said feature pairs by a similarity measure, and optimizing a joint correspondence set of feature pairs by refining the center points to sub-pixel accuracy.

System and method for salient region feature based 3D multi modality registration of medical images

A method of aligning a pair of images with a first image and a second image, wherein said images comprise a plurality of intensities corresponding to a domain of points in a D-dimensional space includes identifying feature points on both images using the same criteria, computing a feature vector for each feature point, measuring a feature dissimilarity for each pair of feature vectors, wherein a first feature vector of each pair is associated with a first feature point on the first image, and a second feature vector of each pair is associated with a second feature point on the second image. A correspondence mapping for each pair of feature points is determined using the feature dissimilarity associated with each feature point pair, and an image transformation is defined to align the second image with the first image using one or more pairs of feature points that are least dissimilar.

Method and system for hybrid rigid registration of 2D/3D medical images

A method of aligning a pair of images includes providing a pair of images with a first image and a second image, wherein the images comprise a plurality of intensities corresponding to a domain of points in a D-dimensional space. Salient feature regions are identified in both the first image and the second image, a correspondence between each pair of salient feature regions is hypothesized, wherein a first region of each pair is on the first image and a second region of each pair is on the second image, the likelihood of the hypothesized correspondence of each pair of feature regions is measured, and a joint correspondence is determined from a set of pairs of feature regions with the greatest likelihood of correspondence.

Growing community of inventors

Lawrenceville, NJ, United States of America

Yiyong Sun