Abstract: Example systems and methods for lesion detection are described herein. An example system includes at least one processor and a memory operably coupled to the at least one processor. The system also includes a candidate selection module configured to receive an image, determine a plurality of candidate points in the image, and select a respective volumetric region centered by each of the candidate points. A portion of a lesion has a high probability of being determined as a candidate point. The system further includes a deep learning network configured to receive the respective volumetric regions selected by the candidate selection module, and determine a respective probability of each respective volumetric region to contain the lesion. Additionally, example methods for training a deep learning network to detect lesions are described herein.

Abstract: Described herein are methods for generating and using a constrained ensemble of GANs. The constrained ensemble of GANs can be used to generate synthetic data that is (1) representative of the original data, and (2) not closely resembling the original data. An example method includes generating a constrained ensemble of GANs, where the constrained ensemble of GANs includes a plurality of ensemble members. The method also includes analyzing performance of the constrained ensemble of GANs by comparing a temporary performance metric to a baseline performance metric, and halting generation of the constrained ensemble of GANs in response to the analysis. The method also includes generating a synthetic dataset using the constrained ensemble of GANs. The synthetic dataset is sufficiently similar to the original dataset to permit data sharing for research purposes but alleviates privacy concerns due to differences in mutual information between synthetic and real data.

Abstract: The invention comprises a sensor unit, and method and system for cargo and personnel detection and tracking. The method comprising the following steps: detecting, or creating for new CCUs, a unique CCU signature, wherein the unique CCU signature is constructed by analyzing images of the CCU capture by a sensor device, and analyzing the images according to one or more of predefined detection methods, and the analysis providing a combination of one or more descriptors defined by CCU feature points, contours, dimensions, weight, colour, movement pattern, neighbour CCU feature point, planned travel route or last known location.

Abstract: A method, system, computer program product, and computer readable media for a semi-automated surface extraction approach to delineating an object of interest (OOI) from 3-D medical image data sets. This approach is imaging modality independent and results in enhanced displays of blob-like anatomies, including internal organs (e.g., cardiac chambers, liver) or disease processes (e.g., tumor masses). In an embodiment of the method: (I) the user provides multiple representative points located on the surface of the OOI using a Multi-Planar Reconstruction (MPR) tool; (2) those Cartesian points are translated into polar coordinates (each 3-D point is represented via two angles and a height), which uniquely define the surface points for a radial object, (3) a Radial-Basis Function (RBF) interpolator, with a Thin Plate Spline (TPS) radial function, finds the height function for the polar domain, and (4) polar domain representation of the OOI surface is converted back to Cartesian coordinates.

Abstract: A method, system, computer program product, and computer readable media for a semi-automated surface extraction approach to delineating an object of interest (OOI) from 3-D medical image data sets. This approach is imaging modality independent and results in enhanced displays of blob-like anatomies, including internal organs (e.g., cardiac chambers, liver) or disease processes (e.g., tumor masses). In an embodiment of the method: (I) the user provides multiple representative points located on the surface of the OOI using a Multi-Planar Reconstruction (MPR) tool; (2) those Cartesian points are translated into polar coordinates (each 3-D point is represented via two angles and a height), which uniquely define the surface points for a radial object, (3) a Radial-Basis Function (RBF) interpolator, with a Thin Plate Spline (TPS) radial function, finds the height function for the polar domain, and (4) polar domain representation of the OOI surface is converted back to Cartesian coordinates.

Abstract: A method for segmenting tubular structures in digital medical images includes extracting a subregion from a 3-dimensional (3D) digital medical image volume containing a vessel of interest, identifying potential vessel centerpoints for each voxel in the subregion by attaching to each voxel a tip of a 3D cone that is oriented in the direction of the voxel's image gradient and having each voxel within the cone vote for those voxels most likely to belong to a vessel centerline, selecting candidates for a second vote image that are both popular according to a first vote image, as well as being consistently voted upon by a radius image, reconfiguring the subregion as a graph where each voxel is represented by a node that is connected to 26 nearest neighbors by n-link edges, and applying a min-cut algorithm to segment the vessel within the subregion.

Abstract: A system and method for tracking and classifying the left ventricle of the heart using cine-delayed enhancement magnetic resonance (Cine-DEMR) are provided. The method for tracking the left ventricle comprises: delineating myocardial borders of the left ventricle in an image of a first phase of a cardiac cycle of the heart; registering the image of the first phase with an image of a second phase of the cardiac cycle; copying the myocardial borders from the first phase onto the second phase; fitting the myocardial borders of the first phase to myocardial borders of the second phase; and refining the myocardial borders of the second phase.

Abstract: A method for segmenting tubular structures in digital medical images includes extracting a subregion from a 3-dimensional (3D) digital image volume containing a vessel of interest, identifying potential vessel centerpoints for each voxel in the subregion by attaching to each voxel a tip of a 3D cone that is oriented in the direction of the voxel's image gradient and having each voxel within the cone vote for those voxels most likely to belong to a vessel centerline, selecting candidates for a second vote image that are both popular according to a first vote image, as well as being consistently voted upon by a radius image, reconfiguring the subregion as a graph where each voxel is represented by a node that is connected to 26 nearest neighbors by n-link edges, and applying a min-cut algorithm to segment the vessel within the subregion.

Abstract: A method and device for image processing to determine image characteristics includes operations of segmenting of first images to create segmentation contours corresponding to the first images; and registering the segmented first images to second images. Further, operations of transferring the segmentation contours from the first images to second images; and fitting the transferred segmentation contours in the second images are performed.

Abstract: A system and method for tracking and classifying the left ventricle of the heart using cine-delayed enhancement magnetic resonance (Cine-DEMR) are provided. The method for tracking the left ventricle comprises: delineating myocardial borders of the left ventricle in an image of a first phase of a cardiac cycle of the heart; registering the image of the first phase with an image of a second phase of the cardiac cycle; copying the myocardial borders from the first phase onto the second phase; fitting the myocardial borders of the first phase to myocardial borders of the second phase; and refining the myocardial borders of the second phase.

Abstract: A method and device for image processing to determine image characteristics includes operations of segmenting of first images to create segmentation contours corresponding to the first images; and registering the segmented first images to second images. Further, operations of transferring the segmentation contours from the first images to second images; and fitting the transferred segmentation contours in the second images are performed.