Abstract: A system and method for characterizing a point in a three-dimensional digital medical image includes determining third order derivatives of the three-dimensional digital medical image at one or more points in the three-dimensional digital medical image. At each of the one or more points, one or more invariant or semi-invariant features that characterize a local geometry at the point is determined using the third order derivatives of the three-dimensional digital medical image. The invariant or semi-invariant features are used for automatic detection of lesions, abnormalities, and other anatomical structures of interest in the three-dimensional digital medical image.

Abstract: A segmentation method comprises clustering spatial, intensity and volumetric shape index to automatically segment a medical lesion. The algorithm has the following steps: (1) calculating volumetric shape index (SI) for each voxel in the image; (2) combining the SI features with the intensity range and the spatial position (x, y, z) to form a 5-dimensional feature vector set; (3) grouping the 5-dimensional feature vector set into clusters; (4) employing a modified expectation-maximization algorithm (EM) considering not only spatial but also shape features on an intensity mode map from the clustering algorithm to merge the neighbouring regions or modes. The joint spatial-intensity-shape feature provides rich information for the segmentation of the anatomic structures of interest, such as lesions or tumours.

Abstract: A computer-implemented method of detecting an object in a three-dimensional medical image comprises determining the values of a plurality of features at each voxel in at least a portion of the medical image. Each feature characterises a respective property of the medical image at a particular voxel. The likelihood probability distribution of each feature is calculated based on the values of the features and prior medical knowledge. A probability map is generated by using Bayes' law to combine the likelihood probability distributions, and the probability map is analysed to detect an object.

Abstract: A segmentation method comprises clustering spatial, intensity and volumetric shape index to automatically segment a medical lesion. The algorithm has the following steps: (1) calculating volumetric shape index (SI) for each voxel in the image; (2) combining the SI features with the intensity range and the spatial position (x, y, z) to form a 5-dimensional feature vector set; (3) grouping the 5-dimensional feature vector set into clusters; (4) employing a modified expectation-maximization algorithm (EM) considering not only spatial but also shape features on an intensity mode map from the clustering algorithm to merge the neighbouring regions or modes. The joint spatial-intensity-shape feature provides rich information for the segmentation of the anatomic structures of interest, such as lesions or tumours.

Abstract: A computer-implemented method of detecting an object in a three-dimensional medical image comprises determining the values of a plurality of features at each voxel in at least a portion of the medical image. Each feature characterises a respective property of the medical image at a particular voxel. The likelihood probability distribution of each feature is calculated based on the values of the features and prior medical knowledge. A probability map is generated by using Bayes' law to combine the likelihood probability distributions, and the probability map is analysed to detect an object.

Abstract: A system and method for characterizing a point in a three-dimensional digital medical image includes determining third order derivatives of the three-dimensional digital medical image at one or more points in the three-dimensional digital medical image. At each of the one or more points, one or more invariant or semi-invariant features that characterize a local geometry at the point is determined using the third order derivatives of the three-dimensional digital medical image. The invariant or semi-invariant features are used for automatic detection of lesions, abnormalities, and other anatomical structures of interest in the three-dimensional digital medical image.

Abstract: A method of operating a data processing system to generate a three-dimensional model of a space from a plurality of measured images of the space. Each measured image includes a view of the space from a corresponding viewpoint. The method divides the space into a plurality of voxels, each voxel being characterized by a location and a color or an indication that the voxel is clear. The method defines a reconstruction of the space by assigning colors and clear values to a set of the voxels. The set is based on at least one linked list of voxels. The reconstruction is characterized by an error value related to the difference between each of the measured images and an image that would be produced by the set of voxels from the corresponding viewpoint as that of the measured image. The colors and clear values are chosen to reduce the error value below a cutoff value.

Abstract: A method of operating a data processing system to generate a three-dimensional model of a space from a plurality of measured images of the space. Each measured image includes a view of the space from a corresponding viewpoint. The method divides the space into a plurality of voxels, each voxel being characterized by a location and a color or an indication that the voxel is clear. The method defines a reconstruction of the space by assigning colors and clear values to a set of the voxels. The reconstruction is characterized by an error value related to the difference between each of the measured images and an image that would be produced by the set of voxels from the corresponding view point as that of the measured image. The colors and clear values are chosen to reduce the error value below a cutoff value.