Abstract: A method for associating a computer aided detection/diagnosis result with an image including the steps of providing one or more digitized images, each image comprising a plurality of intensities corresponding to a domain of points on an N-dimensional grid, performing a computer-aided detection/diagnosis of intensity data of a first image, calculating a hash signature of intensity of a second input image, storing said computer-aided diagnosis results, and storing said hash signature, wherein said hash signature verifies said second image when said second image is displayed with said computer-aided diagnosis result.

Abstract: In one aspect of the present invention, a method for calculating a response value at a first voxel indicative of a global shape in an image is provided. The method includes the steps of (a) determining at least one local shape descriptor associated with each of the at least one local shape descriptor; (b) determining a spread function associated with the each of the at least one local shape descriptor; (c) determining second voxels around the first voxel; (d) calculating values for each the at least one local shape descriptor at each of the second voxels; (e) determining a contribution of each of the second voxels at the first voxel based on the spread functions; and (f) using a combination function to combine the contributions to determine the response value indicative of the global shape.

Abstract: A method of visualizing an object in an image includes presenting an image, selecting a point in an object of interest in said image, determining a main orientation of said object of interest, presenting a first visualization of said object of interest, wherein said first visualization has a first display orientation characterized by the direction of a vector normal to the first visualization plane, and selecting a new point as a center of a new visualization and presenting said new visualization, wherein said new visualization has a new display orientation characterized by the direction of a vector normal to the new visualization plane.

Abstract: A method of visualizing an object in an image includes presenting an image, selecting a point in an object of interest in said image, estimating a gradient of the image in a region about the selected point, calculating a structure tensor from the image gradient, analyzing said structure tensor to determine a main orientation of said object of interest, and presenting a visualization of said object of interest based on the main orientation of the object. Various techniques can be used to increase the robustness of the gradient estimation with respect to noise, and to enhance the visualization of the object-of-interest presented to a user.

Abstract: A method of classifying features in digitized images includes providing a plurality of feature points in an n-dimensional space, wherein said feature points have been extracted from a digitized medical image, formulating a support vector machine to classify said feature point into one of two sets, wherein each said feature classification vector is transformed by an adjacency matrix defined by those points that are nearest neighbors of said feature, and solving said support vector machine by a linear optimization algorithm to determine a classifying plane that separates the feature vectors into said two sets.

Abstract: A cardiac view of a medical ultrasound image is automatically identified. By grouping different views into sub-categories, a hierarchal classifier identifies the views. For example, apical views are distinguished from parasternal views. Specific types of apical or parasternal views are identified based on distinguishing between images of the geneses. Different features are used for classifying, such as gradients, functions of the gradients, statistics of an average frame of data from a clip or sequence of frames, or a number of edges along a given direction. The number of features used may be compressed, such as by classifying a plurality of features into a new feature. For example, alpha weights in a model of features and classes are determined and used as features for classification.

Abstract: The invention concerns an image processing device, including an input for receiving a time series of data sets representing comparable volume digital images, each data having a position component and an intensity component, pre-processing means (18) for modifying the data sets so as to obtain images updated in position and intensity, and comparative processing means (20) for examining sets of time series of image elements and to detect therein signs of variations. The processing means (20) include a modeling function (12) for adjusting a parametric model separately on some of the sets of time series of image elements, to obtain pairs of data (image element, time) and a function for statistical analysis (14) of said pairs of data to isolate the pairs of data representing a significant variation.

Abstract: We propose using different classifiers based on the spatial location of the object. The intuitive idea behind this approach is that several classifiers may learn local concepts better than a “universal” classifier that covers the whole feature space. The use of local classifiers ensures that the objects of a particular class have a higher degree of resemblance within that particular class. The use of local classifiers also results in memory, storage and performance improvements, especially when the classifier is kernel-based. As used herein, the term “kernel-based classifier” refers to a classifier where a mapping function (i.e., the kernel) has been used to map the original training data to a higher dimensional space where the classification task may be easier.

Abstract: CAD (computer-aided decision) support systems, methods and tools for medical imaging are provided, which use machine learning classification for automated detection and marking of regions of interest in medical images. Machine learning methods are used for adapting/optimizing a CAD process by seamlessly incorporating physician knowledge into the CAD process using training data that is obtained during routine use of the CAD system.