Abstract: The first objective comparison of automated and human segmentation of magnetic resonance images, or MRI, using a blinded controlled assessment study. Computers connected over a network divide duties including computerized segmenting of the images, manual segmenting of the images, comparison of the computer segmented images and the manually segmented images, and scoring of the images for accuracy. The scores are evaluated to update configuration parameters of a neural network.

Abstract: An image processing device and methods for performing Rotationally Invariant S-transform (RIST) for an image are provided herein. An example method of determining the RIST magnitude at a pixel is provided herein. Further, an example method of determining RIST magnitudes and statistics in a region of interest is provided herein.

Abstract: An image processing device and methods for performing an S-transform (ST) are provided herein. An example method of generating a compressed form of values of a one-dimensional ST for a time series and generating an approximate form of ST is provided herein. Additionally, an example method of determining local spectrum at a pixel is provided herein. Further, an example method of determining ST magnitudes and statistics in a region of interest (ROI) is provided herein.

Abstract: An image processing device and methods for performing Rotationally Invariant S-transform (RIST) for an image are provided herein. An example method of determining the RIST magnitude at a pixel is provided herein. Further, an example method of determining RIST magnitudes and statistics in a region of interest is provided herein.

Abstract: A method for segmenting a volume dataset is provided. During initialization a level set field within a volume dataset is initialized and an initial set of active voxels is determined in dependence upon the initialized level set field. In an iteration process the level set field for the set of active voxels is updated followed by updating of the set of active voxels. The iteration is continued until the number of active voxels is less than a predetermined threshold. Level set segmentation data are then determined in dependence upon the level set field and provided for, for example, graphical display or storage.

Abstract: An image processing device and methods for performing Rotationally Invariant Stransform (RIST) for an image are provided herein. An example method of determining the RIST magnitude at a pixel is provided herein. Further, an example method of determining RIST magnitudes and statistics in a region of interest is provided herein.

Abstract: An image processing device and methods for performing an S-transform (ST) are provided herein. An example method of generating a compressed form of values of a one-dimensional ST for a time series and generating an approximate form of ST is provided herein. Additionally, an example method of determining local spectrum at a pixel is provided herein. Further, an example method of determining ST magnitudes and statistics in a region of interest (ROI) is provided herein.

Abstract: A parallelized sparse field method for segmenting a volume dataset is provided. Using a processor a level set field within a multi-dimensional dataset comprising a plurality of voxels is initialized in parallel operation. The multi-dimensional dataset is indicative of an object. An initial set of active voxels is then generated in parallel operation in dependence upon the initialized level set field. New active voxels are determined in parallel operation. Duplicate active voxels are then removed in parallel operation, followed by compacting the active voxels in parallel operation. The process is repeated until a quantitative stopping criterion is met and level set segmentation data indicative of a feature of the object are determined in dependence upon the level set field.

Abstract: A method for texture characterization is provided. Multi-dimensional spectrum data are determined by transforming multi-dimensional image data into Fourier domain. The multi-dimensional spectrum data are partitioned into a plurality of partitions, wherein each partition is associated with a predetermined set of orthogonal voice frequencies. The partitioned multi-dimensional spectrum data are then transformed into Stockwell domain resulting in discrete orthonormal Stockwell transform data. The discrete orthonormal Stockwell transform data are then processed to determine data associated with image texture, which are indicative of a feature of the object.

Abstract: A method for texture quantification is provided. Data indicative of a complex-valued local spatial frequency distribution in space-frequency domain are determined by processing a multi-dimensional image data set based on a Stockwell transform with a phase term corresponding to spatial locations of respective frequency components being expressed in terms of radial distance and angle of orientation. The complex-valued local spatial frequency distribution is indicative of a feature of the object. Data indicative of a low frequency energy distribution are determined by filtering the data indicative of a complex-valued local spatial frequency distribution using a band-pass filter having a predetermined low frequency energy bandwidth.

Abstract: A parallelized sparse field method for segmenting a volume dataset is provided. Using a processor a level set field within a multi-dimensional dataset comprising a plurality of voxels is initialized in parallel operation. The multi-dimensional dataset is indicative of an object. An initial set of active voxels is then generated in parallel operation in dependence upon the initialized level set field. New active voxels are determined in parallel operation. Duplicate active voxels are then removed in parallel operation, followed by compacting the active voxels in parallel operation. The process is repeated until a quantitative stopping criterion is met and level set segmentation data indicative of a feature of the object are determined in dependence upon the level set field.

Abstract: A method for segmenting a volume dataset is provided. During initialization a level set field within a volume dataset is initialized and an initial set of active voxels is determined in dependence upon the initialized level set field. In an iteration process the level set field for the set of active voxels is updated followed by updating of the set of active voxels. The iteration is continued until the number of active voxels is less than a predetermined threshold. Level set segmentation data are then determined in dependence upon the level set field and provided for, for example, graphical display or storage.

Abstract: A method for texture characterization is provided. Multi-dimensional spectrum data are determined by transforming multi-dimensional image data into Fourier domain. The multi-dimensional spectrum data are partitioned into a plurality of partitions, wherein each partition is associated with a predetermined set of orthogonal voice frequencies. The partitioned multi-dimensional spectrum data are then transformed into Stockwell domain resulting in discrete orthonormal Stockwell transform data. The discrete orthonormal Stockwell transform data are then processed to determine data associated with image texture, which are indicative of a feature of the object.

Abstract: A method for texture quantification is provided. Data indicative of a complex-valued local spatial frequency distribution in space-frequency domain are determined by processing a multi-dimensional image data set based on a Stockwell transform with a phase term corresponding to spatial locations of respective frequency components being expressed in terms of radial distance and angle of orientation. The complex-valued local spatial frequency distribution is indicative of a feature of the object. Data indicative of a low frequency energy distribution are determined by filtering the data indicative of a complex-valued local spatial frequency distribution using a band-pass filter having a predetermined low frequency energy bandwidth.