Abstract: Systems and methods are provided for aiding the detection and diagnosis of critical heart defects during fetal ultrasound examinations, in which motion video clips are analyzed with machine learning algorithms to identify and select within the motion video clips image frames that correspond to standard views recommended by fetal ultrasound guidelines, and selected image frames are analyzed with machine learning algorithms to detecting and identify morphological abnormalities indicative of critical CHDs associated with the standard views. The results of the analyses are presented for review to the clinician with an overlay for the selected image frames that identifies the abnormalities with graphical or textual indicia. The overlay further may be annotated by the clinician and stored to create documentary record of the fetal ultrasound examination.

Abstract: Image recognition may include obtaining a first image, segmenting the first image into a plurality of first regions by using a target model, and searching for a target region among bounding boxes in the first image that use points in the first regions as centers. The target region is a bounding box in the first image in which a target object is located. The target model is a pre-trained neural network model configured to recognize from an image, a region in which the target object is located. The target model is obtained through training by using positive samples with a region in which the target object is located marked and negative samples with a region in which a noise is located marked. The target region is marked in the first image to improve accuracy of target object detection in an image.

Abstract: A large number of highly accurate learning images are generated without bias at reduced costs. An information processing method for causing a processor to execute: automatically cropping a region including an object from a material image to generate an automatically cropped image; and performing learning related to detection of the object on the basis of the automatically cropped image, wherein the generating of the automatically cropped image further includes generating the automatically cropped image using an automatic cropping machine that is generated by learning on the basis of manually cropped images obtained by manually cropping a region including the object from the material image.

Abstract: A method of cardiac strain analysis uses displacement encoded magnetic resonance image (MRI) data of a heart of the subject and includes generating a phase image for each frame of the displacement encoded MRI data. Phase images include potentially phase-wrapped measured phase values corresponding to pixels of the frame. A convolutional neural network CNN computes a wrapping label map for the phase image, and the wrapping label map includes a respective number of phase wrap cycles present at each pixel in the phase image. Computing an unwrapped phase image includes adding a respective phase correction to each of the potentially-wrapped measured phase values of the phase image, and the phase correction is based on the number of phase wrap cycles present at each pixel. Computing myocardial strain follows by using the unwrapped phase image for strain analysis of the subject.

Abstract: Embodiments of the invention involve combining data representative of the eye obtained from multiple modalities into a virtual model of the eye. The multiple modalities indicate anatomical, physiological, and/or functional features of the eye. The data from different modalities is registered in order to combine the data into the virtual model. Further embodiments involve analysing eye data, for example in the form of the virtual model, using neural networks to obtain insights about medical conditions of the eye, for example the diagnosis or prognosis of conditions, and/or predicting how the eye will respond to certain treatments.

Abstract: Systems and methods for reconstructing medical images are disclosed. Measurement data from positron emission tomography (PET) data, and measurement data from an anatomy modality, such as magnetic resonance (MR) data or computed tomography (CT) data, is received from an image scanning system. A PET image is generated based on the PET measurement data, and an anatomy image is generated based on the anatomy measurement data. A trained neural network is applied to the PET image and the anatomy image to generate an attenuation map. The neural network may be trained based on anatomy and PET images. In some examples, the trained neural network generates an initial attenuation map based on the anatomy image, registers the initial attenuation map to the PET image, and generates an enhanced attenuation map based on the registration. Further, a corrected image is reconstructed based on the generated attenuation map and the PET image.

Abstract: A method of characterizing a serum or plasma portion of a specimen in a specimen container provides an HILN (hemolysis, icterus, lipemia, normal) determination. Pixel data of an input image of the specimen container is processed by a classification network to identify whether the specimen contains plasma or serum. Pixel data representing a plasma sample are forwarded to a segmentation/classification/regression network trained with plasma samples for HILN determination. Pixel data representing a serum sample are forwarded to a transformation network, wherein the serum sample pixel data is transformed into pixel data that matches pixel data of a corresponding previously-collected plasma sample by changing sample color, contrast, intensity, and/or brightness. The transformed serum sample pixel data are forwarded to the segmentation/classification/regression network for HILN determination. Quality check modules and testing apparatus configured to carry out the method are also described, as are other aspects.

Abstract: A modulated segmentation system can use a modulator network to emphasize spatial prior data of an object to track the object across multiple images. The modulated segmentation system can use a segmentation network that receives spatial prior data as intermediate data that improves segmentation accuracy. The segmentation network can further receive visual guide information from a visual guide network to increase tracking accuracy via segmentation.

Abstract: This invention is related to a method to improve the performance of a deep neural network (10) for the identification of a segmentation target (111) in a medical image (12, 110), comprising the steps of performing n training steps on said deep neural network (10) for the identification of said region of interest on two different representations (13, 14) of the same segmentation target (111), said representations (13,14) being definitions of the same segmentation target (111).

Abstract: A method and system for reducing or removing motion artefacts in magnetic resonance (MR) images, the method including the steps of: receiving a motion corrupted MR image; determining a corrected intensity value for each pixel in the motion corrupted MR image by using a neural network; and generating a motion corrected MR image based on the determined corrected intensity values for the pixels in the motion corrupted MR image.

Abstract: A computer-implemented method and system of digitally segmenting teeth in a digital model comprises generating a panoramic image from a 3D digital model of a patient's dentition, labeling, using a first trained neural network, the panoramic image to provide a labeled panoramic image, mapping the labeled panoramic image to corresponding coarse digital surface triangle labels in the 3D digital model to provide a labeled 3D digital model, and segmenting the labeled 3D digital model to provide a segmented 3D digital model. A computer-implemented method and system of generating a panoramic image comprises determining, using a trained neural network, digital tooth bounding region(s) corresponding to digital teeth from a 2D depth map of a patient's dentition, connecting digital tooth bounding region(s) by a spline, determining sampled digital surface points from the sampled spline points; and determining associated digital surface points corresponding to each sampled digital surface point.

Abstract: A method (100) for generating a textual description of a medical image, comprising: receiving (130) a medical image of an anatomical region, the image comprising one or more abnormalities; segmenting (140) the anatomical region in the received medical image from a remainder of the image; identifying (150) at least one of the one or more abnormalities in the segmented anatomical region; extracting (160) one or more features from the identified abnormality; generating (170), using the extracted features and a trained text generation model, a textual description of the identified abnormality; and reporting (180), via a user interface of the system, the generated textual description of the identified abnormality.

Abstract: A method for designing a drilling template, wherein a dental situation is measured by means of a 3D surface measuring device and a 3D surface model of the dental situation is produced and/or measured by means of an X-ray device or an MRI device, wherein the dental situation is measured and a volume model of the dental situation is produced, the method comprising the steps of: applying an artificial neural network for machine learning (convolutional neural network; CNN) to the 3D surface model of the dental situation and/or the volume model of the dental situation and/or to an initial 3D model of the drilling template; and automatically producing a ready made 3D model of the drilling template.

Abstract: A computer-implemented method for correcting artifacts in computed tomography data is provided. The method includes inputting a sinogram into a trained sinogram correction network, wherein the sinogram is missing a pixel value for at least one pixel. The method also includes processing the sinogram via one or more layers of the trained sinogram correction network, wherein processing the sinogram includes deriving complementary information from the sinogram and estimating the pixel value for the at least one pixel based on the complementary information. The method further includes outputting from the trained sinogram correction network a corrected sinogram having the estimated pixel value.

Abstract: Provided are a computer system for automatically searching for a mental disorder diagnosis protocol and an method thereof that may determine at least one test region to be examined for a predetermined mental disorder diagnosis in a brain image of a patient based on a first artificial neural network, may determine a test process for the mental disorder diagnosis for the patient based on a second artificial neural network, and may provide a test protocol for the mental disorder diagnosis for the patient based on the test region and the test process. The computer system may visualize at least one of a position, a shape, a size, and an importance of the test region in the brain image. The test process may include test order of a plurality of test stages in which the brain image is to be used for the mental disorder diagnosis.

Abstract: Disclosed and described herein are systems and methods of performing computer-aided detection (CAD)/diagnosis (CADx) in medical images and comparing the results of the comparison. Such detection can be used for treatment plans and verification of claims produced by healthcare providers, for the purpose of identifying discrepancies between the two. In particular, embodiments disclosed herein are applied to identifying dental caries (“caries”) in radiographs and comparing them against progress notes, treatment plans, and insurance claims.

Abstract: A framework for multi-scan image processing. A single real anatomic image of a region of interest is first acquired. One or more emission images of the region of interest are also acquired. One or more synthetic anatomic images may be generated based on the one or more emission images. One or more deformable registrations of the real anatomic image to the one or more synthetic anatomic images are performed to generate one or more registered anatomic images. Attenuation correction may then be performed on the one or more emission images using the one or more registered anatomic images to generate one or more attenuation corrected emission images.

Abstract: Technologies for determining the accuracy of three-dimensional models include a device having circuitry to obtain two-dimensional images of an anatomical object (e.g., a bone of a human joint), to obtain a candidate three-dimensional model of the anatomical object, and to produce two-dimensional silhouettes of the candidate three-dimensional model. The circuitry is also to apply an edge detection algorithm to the two-dimensional images to produce corresponding edge images and to compare the two-dimensional silhouettes to the edge images to produce a score indicative of an accuracy of the candidate three-dimensional model.

Abstract: Briefly, a variety of embodiments, including the following, are described: a system embodiment and methods that allow random access to voice messages, in contrast to sequential access in existing system embodiments; a system embodiment and methods that allow for the optional use of voice recognition to enhance usability; and a system embodiment and methods that apply to the area of voicemail.

Abstract: A computer-implemented method for generating an artifact corrected reconstructed contrast image from magnetic resonance imaging (MRI) data is provided. The method includes inputting into a trained deep neural network both a synthesized contrast image derived from multi-delay multi-echo (MDME) scan data or the MDME scan data acquired during a first scan of an object of interest utilizing a MDME sequence and a composite image, wherein the composite image is derived from both the MDME scan data and contrast scan data acquired during a second scan of the object of interest utilizing a contrast MRI sequence. The method also includes utilizing the trained deep neural network to generate the artifact corrected reconstructed contrast image based on both the synthesized contrast image or the MDME scan data and the composite image. The method further includes outputting from the trained deep neural network the artifact corrected reconstructed contrast image.