Abstract: A method for image analysis includes receiving compressed, encrypted image data from an imaging session at a healthcare location, unencrypting the image data, and identifying a characteristic of interest in the compressed image data. The method further includes annotating the compressed image data reflecting the identified characteristic of interest to generate computer annotations and streaming the computer annotations to the healthcare location as the imaging session is in progress.

Abstract: Method and system for image registration or image segmentation. The method includes receiving an image which is to be processed by a first machine-learning model to perform, for example, image registration or segmentation, and using a second machine-learning model to determine if the received image is of a quality suitable for the first machine-learning model to act upon.

Abstract: In a system and method for analyzing images, an input image is provided to a computer and is processed therein with a first deep learning model so as to generate an output result for the input image; and applying a second deep learning model is applied to the input image to generate an output confidence score that is indicative of the reliability of any output result from the first deep learning model for the input image.

Abstract: A method is for controlling a system including an imaging modality. The method includes receiving an image and/or a series of images of a portion of a patient from the imaging modality, the field of view of the image and/or of the series of images covering at least a part of an instrument; determining whether the part of the instrument is located within a defined portion of the field of view and/or in a given state selected from a plurality of first states and/or the part of the instrument executes a given movement selected from a plurality of movements; and generating a control command for performing an action depending on whether the part of the instrument is located within the defined portion of the field of view and/or the part of the instrument is in the given state and/or the part of the instrument executes the given movement.

Abstract: In a method and apparatus for training a computer system for use in classification of an image by processing image data representing the image, image data are compressed and then loaded into a programmable quantum annealing device that includes a Restricted Boltzmann Machine. The Restricted Boltzmann Machine is trained to act as a classifier of image data, thereby providing a trained Restricted Boltzmann Machine; and, the trained Restricted Boltzmann Machine is used to initialize a neural network for image classification thereby providing a trained computer system for use in classification of an image.

Abstract: The disclosure relates to a method of determining a transformation between coordinate frames of sets of image data. The method includes receiving a model of a structure extracted from first source image data, the first source image data being generated according to a first imaging modality and having a first data format, wherein the model has a second data format, different from the first data format. The method also includes determining, using an intelligent agent, a transformation between coordinate frames of the model and first target image data, the first target image data being generated according to a second imaging modality different to the first imaging modality.

Abstract: Method and system for image registration or image segmentation. The method includes receiving an image which is to be processed by a first machine-learning model to perform, for example, image registration or segmentation, and using a second machine-learning model to determine if the received image is of a quality suitable for the first machine-learning model to act upon.

Abstract: A method of training a computer system for use in determining a transformation between coordinate frames of image data representing an imaged subject. The method trains a learning agent according to a machine learning algorithm, to determine a transformation between respective coordinate frames of a number of different views of an anatomical structure simulated using a 3D model. The views are images containing labels. The learning agent includes a domain classifier comprising a feature map generated by the learning agent during the training operation. The classifier is configured to generate a classification output indicating whether image data is synthesized or real images data. Training includes using unlabeled real image data to training the computer system to determine a transformation between a coordinate frame of a synthesized view of the imaged structure and a view of the structure within a real image.

Abstract: A method of extracting mechanical activation of the left ventricle from a sequence of contrasted X-ray fluoroscopy images is provided. The method includes: processing the image sequence to perform segmentation of the coronary veins; annotating branches of the segmented coronary veins; tracking the positions of the annotations throughout the fluoroscopy image sequence; computing the principle components of the motion of the annotations from the tracked positions; projecting the motion of the annotations to the axis corresponding to a first principle component; and analyzing the resulting motion curves to identify a latest activating region of the left ventricle.

Abstract: Techniques are disclosed related to using anatomical mask data acquired via magnetic resonance imaging (MRI) scans to train a convolutional neural network (CNN). The training may include the verification of cardiac scar tissue locations data obtained from the anatomical mask data with a reliable system for doing so, such as ground truth data from enhanced cardiac MRI late gadolinium enhanced (LGE) scans. Once the CNN is adequately trained using the anatomical mask data, the CNN may be used to identify cardiac scar tissue from image data obtained from medical imaging modalities other than MRI.

Abstract: In a system and method for analyzing images, an input image is provided to a computer and is processed therein with a first deep learning model so as to generate an output result for the input image; and applying a second deep learning model is applied to the input image to generate an output confidence score that is indicative of the reliability of any output result from the first deep learning model for the input image.

Abstract: A method of training a computer system for use in determining a transformation between coordinate frames of image data representing an imaged subject. The method trains a learning agent according to a machine learning algorithm, to determine a transformation between respective coordinate frames of a number of different views of an anatomical structure simulated using a 3D model. The views are images containing labels. The learning agent includes a domain classifier comprising a feature map generated by the learning agent during the training operation. The classifier is configured to generate a classification output indicating whether image data is synthesized or real images data. Training includes using unlabeled real image data to training the computer system to determine a transformation between a coordinate frame of a synthesized view of the imaged structure and a view of the structure within a real image.

Abstract: In a method and apparatus for training a computer system for use in classification of an image by processing image data representing the image, image data are compressed and then loaded into a programmable quantum annealing device that includes a Restricted Boltzmann Machine. The Restricted Boltzmann Machine is trained to act as a classifier of image data, thereby providing a trained Restricted Boltzmann Machine; and, the trained Restricted Boltzmann Machine is used to initialize a neural network for image classification thereby providing a trained computer system for use in classification of an image.

Abstract: A method of extracting mechanical activation of the left ventricle from a sequence of contrasted X-ray fluoroscopy images is provided. The method includes: processing the image sequence to perform segmentation of the coronary veins; annotating branches of the segmented coronary veins; tracking the positions of the annotations throughout the fluoroscopy image sequence; computing the principle components of the motion of the annotations from the tracked positions; projecting the motion of the annotations to the axis corresponding to a first principle component; and analyzing the resulting motion curves to identify a latest activating region of the left ventricle.

Abstract: A method for registering image data sets of a target region of a patient includes selecting a first anatomical structure only or at least partially only visible in the first image data set, and a second anatomical structure only or at least partially only visible in the second image data set, such that there is a known geometrical relationship between extended segments of the anatomical structures; automatically determining a first geometry information describing the geometry of at least a part of the first anatomical structure and a second geometry information describing the geometry of at least a part of the second anatomical structure, neither information being sufficient to enable registration of the image data sets on its own; automatically optimizing transformation parameters describing a rigid transformation of one of the anatomical structures with respect to the other and geometrical correspondences; and determining registration information from the optimized transformation parameters.

Abstract: The disclosure relates to a method of determining a transformation between coordinate frames of sets of image data. The method includes receiving a model of a structure extracted from first source image data, the first source image data being generated according to a first imaging modality and having a first data format, wherein the model has a second data format, different from the first data format. The method also includes determining, using an intelligent agent, a transformation between coordinate frames of the model and first target image data, the first target image data being generated according to a second imaging modality different to the first imaging modality.

Abstract: In order to improve depth perception for an image displayed during a laparoscopic surgery, a representation of a shadow of a tool included in the image and used in the laparoscopic surgery is identified and introduced into the image. A processor augments a three-dimensional (3D) model including a 3D representation of a surface of an object included in the image, and a representation of the tool by introducing a virtual light source into the 3D model to generate a virtual shadow within the 3D model. The processor subtracts the representation of the shadow out of the augmented 3D model and superimposes the representation of the shadow on the image to be displayed during the laparoscopic surgery.

Abstract: In order to improve depth perception for an image displayed during a laparoscopic surgery, a representation of a shadow of a tool included in the image and used in the laparoscopic surgery is identified and introduced into the image. A processor augments a three-dimensional (3D) model including a 3D representation of a surface of an object included in the image, and a representation of the tool by introducing a virtual light source into the 3D model to generate a virtual shadow within the 3D model. The processor subtracts the representation of the shadow out of the augmented 3D model and superimposes the representation of the shadow on the image to be displayed during the laparoscopic surgery.

Abstract: A method is for controlling a system including an imaging modality. The method includes receiving an image and/or a series of images of a portion of a patient from the imaging modality, the field of view of the image and/or of the series of images covering at least a part of an instrument; determining whether the part of the instrument is located within a defined portion of the field of view and/or in a given state selected from a plurality of first states and/or the part of the instrument executes a given movement selected from a plurality of movements; and generating a control command for performing an action depending on whether the part of the instrument is located within the defined portion of the field of view and/or the part of the instrument is in the given state and/or the part of the instrument executes the given movement.

Abstract: Systems and methods for model augmentation include receiving intra-operative imaging data of an anatomical object of interest at a deformed state. The intra-operative imaging data is stitched into an intra-operative model of the anatomical object of interest at the deformed state. The intra-operative model of the anatomical object of interest at the deformed state is registered with a pre-operative model of the anatomical object of interest at an initial state by deforming the pre-operative model of the anatomical object of interest at the initial state based on a biomechanical model. Texture information from the intra-operative model of the anatomical object of interest at the deformed state is mapped to the deformed pre-operative model to generate a deformed, texture-mapped pre-operative model of the anatomical object of interest.