Abstract: A method and apparatus for automatically performing medical image analysis tasks using deep image-to-image network (DI2IN) learning. An input medical image of a patient is received. An output image that provides a result of a target medical image analysis task on the input medical image is automatically generated using a trained deep image-to-image network (DI2IN). The trained DI2IN uses a conditional random field (CRF) energy function to estimate the output image based on the input medical image and uses a trained deep learning network to model unary and pairwise terms of the CRF energy function. The DI2IN may be trained on an image with multiple resolutions. The input image may be split into multiple parts and a separate DI2IN may be trained for each part. Furthermore, the multi-scale and multi-part schemes can be combined to train a multi-scale multi-part DI2IN.

Abstract: A block processing procedure to determine coding-block or transform-block sizes for encoded and decoding video data. An encoder obtains unencoded video data and selects a region-of-interest in the video data. Image characteristics of the video data are determined from the region-of-interest and a block size is determined from those image characteristics. The video data is encoded using the block size without storing the block size in a header associated with the encoded data. A decoder obtains the encoded data and selects a second region-of-interest in the encoded data that corresponds to the unencoded data region-of-interest. Second image characteristics are determined from the second region-of-interest, where the second characteristics are the same as the characteristics determined for the unencoded data. A block size is determined from the second image characteristics without using a header associated with the encoded data. The encoded data is decoded based on the determined block size.

Abstract: A method for determining a clinical characteristic of a body vessel segment including providing, to a computing device, a three-dimensional reconstruction of a body vessel containing the body vessel segment. A segmented angiography recording of the body vessel segment is provided to the computing device. The computing device extracts at least one global feature of the body vessel from the three-dimensional reconstruction and extracts at least one local feature of the body vessel segment from the angiography recording. The clinical characteristic is determined for the body vessel segment as a function of the at least one extracted local feature and the at least one extracted global feature.

Abstract: Systems, computer-implemented methods, and computer readable media for generating a synthetic image of an anatomical portion based on an origin image of the anatomical portion acquired by an imaging device using a first imaging modality are disclosed. These systems may be configured to receive the origin image of the anatomical portion acquired by the imaging device using the first imaging modality, receive a convolutional neural network model trained for predicting the synthetic image based on the origin image, and convert the origin image to the synthetic image through the convolutional neural network model. The synthetic image may resemble an imaging of the anatomical portion using a second imaging modality differing from the first imaging modality.

Abstract: A medical image processing apparatus includes an acquisition unit and a processing unit. The acquisition unit acquires first and second medical images in time series. The processing unit calculates a first parameter indicating a transition of luminance at a first point based on the luminance at the first point on the first medical image and luminance at a second point on the second medical image, calculates a second parameter indicating a transition of luminance at a third point based on the luminance at the third point on the first medical image and luminance at a fourth point on the second medical image, visualizes the first medical image by superimposing the first parameter on the first point and superimposing the second parameter on the third point; and visualizes the second medical image by superimposing the first parameter on the second point and superimposing the second parameter on the fourth point.

Abstract: The present invention relates to an apparatus for providing a synthetic representation of a vascular structure. It is described to provide (210) at least one 2D X-ray image comprising 2D X-ray image data of a vascular structure of a patient's body part is provided. A 3D model of the body part is provided (220), the 3D model comprising a 3D modelled vascular structure. A 2D projection of the 3D model of the body part is determined (230), the 2D projection of the 3D model of the body part comprising a 2D projection of the 3D modelled vascular structure. The 3D model of the body part is transformed (240). The transform of the 3D model of the body part comprises a determination of the pose of the 3D model of the body part such that a 2D projection of the 3D modelled vascular structure associated with the transformed 3D model of the body part is representative of the 2D X-ray image data of the vascular structure of the patient's body part.

Abstract: A method and system are provided for obtaining a true shape of objects in a medical image. The method includes receiving X-ray stand geometry parameter values, an Anglepoint, and an Anglestand values; selecting a plurality of pixels representing an object in the X-ray image; and determining magnification of each pixel in the X-ray image, that enables accurate determination of size and shape of imaged objects. The method may include determining an actual size of the object in the X-ray image and selectively reshaping the object in the X-ray image to obtain a true shape of the object.

Abstract: An ultrasound imaging system includes a beamformer configured to generate 2-D images offset from each other based on sweeping motion of a transducer array during a 3-D ultrasound imaging procedure producing volumetric data. The ultrasound imaging system further includes a 2-D mask processor configured to generate a 2-D mask image for each of the 2-D images. Each of the 2-D mask images includes a contour of a predetermined structure of interest in the volumetric data. The ultrasound imaging system further includes a 3-D mask processor configured to segment the structure from the 3-D image with the 2-D mask images, generating a 3-D segmentation.

Abstract: A system for determining a region of interest in an image. The system includes a memory and an electronic processor. The electronic processor included in the system is connected to the memory and is configured to initialize internal states of nodes of a spatial lattice. Each node of the spatial lattice corresponds to a pixel of the image and is connected to at least one node representing a neighboring pixel of the image. The electronic processor is also configured to iteratively update, using a neural network, the internal states of each nodes in the spatial lattice using spatially gated propagation and identify the region of interest within the image based on the internal states of the nodes at a convergence of the spatial lattice.

Abstract: Disclosed herein are system, method, and computer program product embodiments providing color theme maintenance for presentations. An embodiment operates by receiving a border image, a background color rule, a text color rule, and compiling the rules into a theme for a presentation. The compiled theme is provided to an application configured to display the presentation including the compiled theme in accordance with the background color rule and the text color rule.

Abstract: Provided is a technique for deriving a mathematical function for defining arterial and venous blood flow in the body by using a four-dimensional medical image. A method of analyzing blood flow by using a medical image according to an embodiment of the present disclosure includes: determining a position of a blood vessel from four-dimensional medical image data that is obtained by combining data of three-dimensional medical images of a patient's body captured at a preset period; deriving a primary function for an arterial input function and a venous output function by using a vascular signal in a head region and a vascular signal in a heart region from among vascular signals included in three-dimensional medical image data for the position of the blood vessel determined in the determining; and deriving a secondary function that is a final function for the arterial input function and the venous output function by using the primary function and a vascular signal in a neck region.

Abstract: A radiation image processing device which performs image processing to a moving image that is obtained by emitting radiation to a subject, the radiation image processing device including a hardware processor that: extracts a reference region from each of a plurality of frame images which form the moving image; calculates an image processing condition of the reference region for each of the frame images from which the reference region is extracted; determines a standard image processing condition which is a standard based on the calculated image processing condition of each of the frame images; and performs image processing by applying the determined standard image processing condition to each of the frame images.

Abstract: A convolutional neural network (CNN) is applied to identifying tumors in a histological image. The CNN has one channel assigned to each of a plurality of tissue classes that are to be identified, there being at least one class for each of non-tumorous and tumorous tissue types. Multi-stage convolution is performed on image patches extracted from the histological image followed by multi-stage transpose convolution to recover a layer matched in size to the input image patch. The output image patch thus has a one-to-one pixel-to-pixel correspondence with the input image patch such that each pixel in the output image patch has assigned to it one of the multiple available classes. The output image patches are then assembled into a probability map that can be co-rendered with the histological image either alongside it or over it as an overlay. The probability map can then be stored linked to the histological image.

Abstract: A magnetic resonance (MR) imaging system includes a memory for storing machine executable instructions and for storing pulse sequence commands to acquire the measured MR data according to a compressed sensing MR imaging protocol; and a processor for controlling the system. The MR imaging system with the pulse sequence commands acquires the measured MR data; reconstruct an intermediate MR image according to the compressed sensing MR imaging protocol; calculate a predicted data portion for each of the measured data portions; calculate a residual for each of the measured data portions; identify one or more of the measured data portions as outlier data portions; and reconstruct a corrected MR image according to the compressed sensing MR imaging protocol, wherein the one or more outlier data portions are excluded from the reconstruction of the corrected MR image.

Abstract: A system and a method are disclosed that allow for generation of a model or reconstruction of a model of a subject based upon acquired image data. The image data can be acquired in a substantially mobile system that can be moved relative to a subject to allow for image acquisition from a plurality of orientations relative to the subject. The plurality of orientations can include a first and final orientation and a predetermined path along which an image data collector or detector can move to acquire an appropriate image data set to allow for the model of construction.

Abstract: A method and apparatus for deep learning based automatic bone removal in medical images, such as computed tomography angiography (CTA) volumes, is disclosed. Bone structures are segmented in a 3D medical image of a patient by classifying voxels of the 3D medical image as bone or non-bone voxels using a deep neural network trained for bone segmentation. A 3D visualization of non-bone structures in the 3D medical image is generated by removing voxels classified as bone voxels from a 3D visualization of the 3D medical image.

Abstract: A MRI device includes: structure applying a main magnetic field on an axis Z over a sample zone; structure emitting a magnetic field gradient and structure emitting a radiofrequency pulse, and a controller. The controller performs on the sample zone, a sequence including: a radiofrequency pulse and/or phase at each repetition; and a spatial gradient of the component along the Z axis. The controller is programmed so that, in the course of repeated applications of the pulse and of the gradient of the sequence of one and the same set: the radio frequency pulse follows, between its repeated applications, a periodic series for its amplitude and for a series U+t=v+i?v; and each repeated application of the gradient of magnetic field of the sequence a, according to a coding spatial direction, a non zero timing integral equal to A and identical for each application of gradient of this set.

Abstract: Deep reinforcement machine learning is used to control denoising (e.g., image regularizer) in iterative reconstruction for MRI compressed sensing. Rather than requiring different machine-learnt networks for different scan settings (e.g., acceleration of the MR compressed sensing), reinforcement learning creates a policy of actions to provide denoising and data fitting through iterations of the reconstruction given a range of different scan settings. This allows a user to scan as appropriate for the patient, the MR system, the application, and/or preferences while still providing an optimized reconstruction under sampling resulting from the MR compressed sensing.

Abstract: A method for dual-contrast unenhanced magnetic resonance angiography includes iteratively acquiring flow-dependent slices and flow-independent slices in a region. Each iteration of the acquisition process comprises identifying a flow-dependent slice location within the region and identifying a flow-independent slice location upstream from the flow-dependent slice location according to blood flow in the region. Each iteration further includes applying a first radio frequency (RF) saturation pulse to the region such that MR signals from veins in the region are substantially suppressed, and applying a second RF saturation pulse to the flow-dependent slice location such that MR signals from background muscle and arterial blood in the region are substantially suppressed. A flow independent slice is acquired at the flow-independent slice location after the second RF saturation pulse is applied and before unsaturated arterial blood has maximally flowed into the region.

Abstract: An apparatus for illuminating or masking an object and a method of using same. The apparatus includes a spatial light modulator transmitting, a structured pulsed laser beam from a pupil plane to at least one image plane in a field of view. The apparatus further includes a lidar detector receiving reflected laser beam reflected from the at least one image plane. For example, the lidar detector detects range, position, and/or time data for at least one object of interest or at least one object of disinterest. Using the detected data, the spatial light modulator illuminates object of interest or masks an object of disinterest, depending on a user's application.

Abstract: Embodiments of an automated method of processing fingerprint images, identity information is extracted from prints typically classified as having “no identification value” because of sparse or missing minutiae by capturing ridge contour information. Bezier approximations of ridge curvature are used as Ridge Specific Markers. Control points arising from Bezier curves generate unique polygons that represent the actual curve in the fingerprint. The Bezier-based descriptors are then grouped together and compared to corresponding reference print Ridge Specific Marker data. The method makes it possible to fuse a plurality of individual latent print portions into a single descriptor of identity and use the resulting data for comparison and identification. Processing of poor quality reference prints according to the methods disclosed renders these prints useable for reference purposes.

Abstract: The invention relates to an imaging device (10) for registration of different imaging modalities, an imaging system (1) for registration of different imaging modalities, a method for registration of different imaging modalities, a computer program element for controlling such device and a computer readable medium having stored such computer program element. The imaging system (1) for registration of different imaging modalities comprises a first imaging modality (2), a second imaging modality (3), and an imaging device (10) for registration of different imaging modalities. The imaging device (10) for registration of different imaging modalities comprises a model provision unit (11), a first image data provision unit (12), a processing unit (13), a second image data provision unit (14), and a registration unit (15). The model provision unit (11) provides a cavity model of a visceral cavity.

Abstract: An X-ray diagnostic apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to measure respective profiles on contrast media concentration in regions of interest including blood vessels set at about the same position in two subtraction images of subject's head taken from about the same direction at different radiography times. The processing circuitry is configured to determine a correction factor so that the two profiles measured are approximately matched. The processing circuitry is configured to correct at least one of the two subtraction images on the basis of the correction factor determined. The processing circuitry is configured to control so as to display information based on the two subtraction images that at least one thereof has been corrected on a display.

Abstract: A medical image processing apparatus according to an embodiment includes an estimation circuitry and a tracking circuitry. The estimation circuitry is configured to estimate the activity of the myocardium across a plurality of images at different time phases from a group of images where a plurality of images containing a myocardium are chronologically arranged. The tracking circuitry is configured to set a search range for tracking the myocardium in the group according to the activity of the myocardium and perform the tracking.

Abstract: A method for reconstructing an image data set from magnetic resonance data is provided. First measurement data is captured using an image capturing device. The first measurement data is captured using temporal and/or spatial subsampling and is used for reconstructing the image data set with a compressed sensing algorithm in which a boundary condition that provided agreement with the measurement data and a target function that is used in an iterative optimization. The compressed sensing algorithm evaluates candidate data sets for the image data set are used. In the reconstruction using the compressed sensing algorithm, in addition to the first measurement data, second measurement data that is captured by a second imaging modality that is different from the first imaging modality of the first measurement data but by the same image capturing device. The second measurement data is registered to the first measurement data, by a modification of the boundary condition and/or target function.

Abstract: A system for parameterized FPGA (Field Programable Gate Array) implementation of real-time SENSE (SENSitivity Encoding) reconstruction including: a sensitivity maps memory configured to store sensitivity map data; an aliased image memory configured to store aliased image data acquired from a scanner; a reconstructed image memory configured to store reconstructed image data; a parameterized complex matrix multiplier; a pseudo-inverse calculator; a magnitude image block; and a controller; wherein sensitivity map data from the sensitivity maps memory is transferred to the pseudo-inverse calculator; wherein data from the pseudo-inverse calculator and the aliased image data from the aliased image memory is transferred to the complex matrix multiplier; wherein data from the complex matrix multiplier is transferred to the magnitude image block; wherein the controller is configured to generate an address of the sensitivity map memory and an address of the aliased image memory to access the encoding matrix and corresp

Abstract: Aspects of the disclosure provide a method for denoising a reconstructed picture. The method can include receiving reconstructed video data corresponding to a picture, dividing the picture into current patches, forming patch groups each including a current patch and a number of reference patches that are similar to the current patch, denoising the patch groups to modify pixel values of the patch groups to create a filtered picture, and generating a reference picture based on the filtered picture for encoding or decoding a picture. The operation of denoising the patch groups includes deriving a variance of compression noise in the respective patch group based on a compression noise model. A selection of model parameters is determined based on coding unit level information.

Abstract: In part, the disclosure relates to shadow detection and shadow validation relative to data sets obtained from an intravascular imaging data collection session. The methods can use locally adaptive thresholds and scan line level analysis relative to candidate shadow regions to determine a set of candidate shadows for validation or rejection. In one embodiment, the shadows are stent strut shadows, guidewire shadows, side branch shadows or other shadows.

Abstract: A system and method for reconstructing an image of a subject includes acquiring a reference dataset and reconstructing a prior image of the subject from the reference dataset and selecting at least a portion of the prior image that corresponds to a portion of the prior image of the subject that is free of artifacts. The method also includes acquiring a medical imaging dataset of the subject, performing a vertical comparison of the medical imaging dataset and the reference dataset to create a data inconsistency metric, and repeating the preceding steps to create a plurality of data inconsistency metrics. The method further includes performing a horizontal comparison of the data inconsistency metrics to identify inconsistent data, compensating for the inconsistent data, and reconstructing an image of the subject with reduced artifacts compared to the prior image.

Abstract: A method of analyzing and assessing an image comprises the steps of: (a) obtaining said image in a digital form; (b) processing said image; and (c) establishing at least one edge line within said image. The step of processing said image further includes: (a) building a 3D graph of intensity distribution within said image; (b) calculating identifying vector angle values (IVAVs); said angles adjoin to points of said 3D intensity distribution graph which correspond to individual pixels of said image; said angles lie in a plane defined by an intensity axis of said image; (c) setting a range of lower and upper thresh-hold level values (THLV1 and THLV2) between which edge points are indicatable; and (d) indicating edge points according to said THLV1 and THLV2.

Abstract: A reconstruction image is generated with a small number of updates, with the use of an iterative approximation method. A specified tomographic image of a subject is received, and a process is performed two or more times, where an update process is performed according to the iterative approximation method, using the tomographic image as an initial image and an update image is obtained. Then, an update vector corresponding to a difference between thus generated update images of the update process performed twice is multiplied by predetermined coefficients, so as to generate an estimated update vector. Using this vector, an update image is generated. Then, this update image is used as a new initial image, and a process is repeated where the update process is performed according to the iterative approximation method and an update image is obtained, thereby generating a tomographic image of the subject.

Abstract: As set forth herein low energy signal data and high energy signal data can be acquired. A first material decomposed (MD) image of a first material basis and a second material decomposed (MD) image of a second material basis can be obtained using the low energy signal data and the high energy signal data. At least one of the first or second MD image can be input into a guide filter for output of at least one noise reduced and cross-contamination reduced image. A computed tomography (CT) imaging system can be provided that includes an X-ray source and a detector having a plurality of detector elements that detect X-ray beams emitted from the X-ray source. Low energy signal data and high energy signal data can be acquired using the detector.

Abstract: A method and a wearable apparatus for disease diagnosis are provided. The method is applied to the wearable apparatus with an image capturing unit and a display unit. In this method, a plurality of input images in a field of view of the wearable apparatus are captured by using the image capturing unit, wherein each of the input images contains an array of pixels. The variations of the pixel values in a time domain are analyzed. The pixel variations within a specific frequency range are magnified and the magnified pixel variations are added onto the original ones to generate an output image. The output image is overlapped with a current image in the field of view of the wearable apparatus and displayed on the display unit.

Abstract: A method and a device for filtering the aberrations of disparity or depth images using an adaptive approach are described. The method allows the local filtering of those points which are not spatially coherent in their 3D neighborhood, according to a criterion derived from a geometrical reality of the transformations carried out on the light signals. Advantageously, the noise filtering method may be applied to a dense depth image or to a dense disparity image.

Abstract: Provided are systems and methods for analyzing images. An exemplary method can comprise receiving at least one image having one or more annotations indicating a feature. The method can comprise generating training images from the at least one image. Each training image can be based on a respective section of the at least one image. The training images can comprise positive images having the feature and negative images without the feature. The method can comprise generating a feature space based on the positive images and the negative images. The method can further comprise identifying the feature in one or more unclassified images based upon the feature space.

Abstract: A system and method of computer-aided detection (CAD or CADe) of medical images that utilizes persistence between images of a sequence to identify regions of interest detected with low interference from artifacts to reduce false positives and improve probability of detection of true lesions, thereby providing improved performance over static CADe methods for automatic ROI lesion detection.

Abstract: A system and method includes reception of a plurality of fill frames of a patient volume, each of the plurality of fill frames depicting a contrast medium within the patient volume at a respective time, identification, for each pixel location of the fill frames, of a fill frame whose pixel at the pixel location is associated with a pixel value which represents a greater level of contrast medium than the pixel values of pixels at the pixel location within the others of the plurality of fill frames, generation of a peak contrast fill frame corresponding to each fill frame, the peak contrast fill frame corresponding to a given fill frame including, at pixel locations for which the given fill frame was identified, pixels associated with pixel values of the given fill frame, and storage of the plurality of peak contrast fill frames.

Abstract: This invention provides a technique of appropriately managing a high-resolution tomographic image and effectively displaying an arbitrary cross section even under an environment where a 3D texture has a size limitation. An information processing apparatus includes a determination unit adapted to determine whether a size of a plurality of tomographic images acquired from a single object is not more than a predetermined size, a management unit adapted to, upon determining that the size is not more than the predetermined size, manage the plurality of tomographic images as three-dimensional voxel data, a decision unit adapted to decide a cross-sectional image as a display target of an object image managed as the three-dimensional voxel data, and a display control unit adapted to cause a display unit to display the decided cross-sectional image.

Abstract: A system and method are presented for neural network based feature extraction for acoustic model development. A neural network may be used to extract acoustic features from raw MFCCs or the spectrum, which are then used for training acoustic models for speech recognition systems. Feature extraction may be performed by optimizing a cost function used in linear discriminant analysis. General non-linear functions generated by the neural network are used for feature extraction. The transformation may be performed using a cost function from linear discriminant analysis methods which perform linear operations on the MFCCs and generate lower dimensional features for speech recognition. The extracted acoustic features may then be used for training acoustic models for speech recognition systems.

Abstract: Methods and systems for visualizing, simulating, modifying and/or 3D printing objects are provided. The system is an end-to-end system that can take as input 3D objects and allow for various levels of user intervention to produce the desired results.

Abstract: Digital subtraction angiography is used to improve the contrast of images of patient vasculature. A non-contrast image is recorded with no contrast medium injected into the patient, and then a succession of contrast images is captured after the injection of a contrast medium. The non-contrast image is successively registered to the contrast images, and then subtracted. This removes background structures, but leaves the vasculature untouched, making blood vessels easier to see. Artifacts can remain when different motion layers are present in the X-ray image (for example, the spine and the lungs). This application discusses a technique to prevent artifacts occurring when different motion layers are present in an X-ray frame or sequence.

Abstract: A method calculates a four-dimensional DSA dataset from x-ray datasets. Each of the x-ray datasets contains a two-dimensional x-ray projection of an examination volume in relation to a direction of projection and a recording time. A first three-dimensional DSA dataset of a first reconstruction volume is determined based on the x-ray datasets. The first reconstruction volume is a part of the examination volume. A second three-dimensional DSA dataset of a second reconstruction volume is determined based on the x-ray datasets. The second reconstruction volume is a part of the first reconstruction volume. The second three-dimensional DSA dataset is segmented. The x-ray datasets are normalized based on the first three-dimensional DSA dataset. A four-dimensional DSA dataset is calculated by back projection of the normalized x-ray datasets onto the segmented second three-dimensional DSA dataset.

Abstract: An image processing apparatus includes a structure reducing unit which reduces spatial signal change by a predetermined structure in an X-ray image group in which a plurality of images obtained by capturing a same site of a same subject a plurality of times using an X-ray are aligned in a time series. The structure reducing unit reduces the spatial signal change by the predetermined structure in each image of the X-ray image group using information of an image different from the image of the X-ray image group being processed.

Abstract: A method is provided for representing a first structure of a body region by digital subtraction angiography. The method includes: receiving a filler image of the body region created by an angiography apparatus, which represents a second structure of the body region and the first structure with a first contrast medium concentration in the first structure; determining a mask image of the body region representing the second structure; determining a subtraction image by editing out of the second structure from the filler image by the mask image; determining a guidance image representing the first structure based on the subtraction image; reducing image noise of the subtraction image by the guidance image; and representing the first structure based on the noise-reduced subtraction image.

Abstract: A device and method for visualization of the intravascular creation of autologous valves, and particularly venous valve, is disclosed herein. One aspect of the present technology, for example, is directed toward a delivery catheter that can include a lumen configured to receive a dissection assembly and a trough having a plurality of transducers electrically coupled to a proximal portion of the delivery catheter. At least one of the transducers can be configured to emit a signal towards a portion of a blood vessel adjacent the trough, and at least one of the transducers can be configured to receive a reflection of the emitted signal.

Abstract: An apparatus includes an interface unit configured to provide, to a terminal, an interface supporting one or more modes, and display, on the interface, an image including a contour of a region of interest. The apparatus further includes a contour modification unit configured to modify the contour based on a mode selected by a user from the one or more modes, and an operation performed by the user.

Abstract: The disclosure relates generally to the field of vascular system and peripheral vascular system data collection, imaging, image processing and feature detection relating thereto. In part, the disclosure more specifically relates to methods for detecting position and size of contrast cloud in an x-ray image including with respect to a sequence of x-ray images during intravascular imaging. Methods of detecting and extracting metallic wires from x-ray images are also described herein such as guidewires used in coronary procedures. Further, methods for of registering vascular trees for one or more images, such as in sequences of x-ray images, are disclosed. In part, the disclosure relates to processing, tracking and registering angiography images and elements in such images. The registration can be performed relative to images from an intravascular imaging modality such as, for example, optical coherence tomography (OCT) or intravascular ultrasound (IVUS).

Abstract: Methods and systems for automatically determining a clinical image or portion thereof for display to a diagnosing physician. One system includes an electronic processor and an interface for communicating with at least one data source. The electronic processor is configured to receive training information from the at least one data source and determine a subset of images included in each of the plurality of image studies displayed to one or more diagnosing physicians. The electronic processor is also configured to perform machine learning to develop a model based on the training information and the subset of images included in each of the plurality of image studies and receive the image study. The electronic processor is also configured to process the image study using the model to determine a subset of the plurality of images and flag the subset of the plurality of images for manual review by the diagnosing physician.

Abstract: A method and system for multi-scale anatomical and functional modeling of coronary circulation is disclosed. A patient-specific anatomical model of coronary arteries and the heart is generated from medical image data of a patient. A multi-scale functional model of coronary circulation is generated based on the patient-specific anatomical model. Blood flow is simulated in at least one stenosis region of at least one coronary artery using the multi-scale function model of coronary circulation. Hemodynamic quantities, such as fractional flow reserve (FFR), are computed to determine a functional assessment of the stenosis, and virtual intervention simulations are performed using the multi-scale function model of coronary circulation for decision support and intervention planning.

Abstract: A method is for determining a transformation for image registration of a first image relative to a second image. The method includes ascertaining a test series of test elements including a test transformation and a test value, the ascertaining including ascertaining the test transformation based on a sequence of test transformations and/or based on previously ascertained test elements, transforming the first image via the ascertained test transformation, ascertaining a difference image, and ascertaining the test value of the test element based on the difference image such that the test value is a measure for an extension of a frequency distribution of values of pixels of the difference image in a direction of pixel value increase. It further includes determining a minimum test value based on test values encompassed by the test elements and determining the transformation which is the test transformation of a test element including the minimum test value.