Abstract: A method and apparatus for extracting centerline representations of vascular structures in medical images is disclosed. A vessel orientation tensor for each of a plurality of voxels associated with the target vessel, such as a coronary artery, in a medical image, such as a coronary tomography angiography (CTA) image, using a trained vessel orientation tensor classifier. A flow field is estimated for the plurality of voxels associated with the target vessel in the medical image based on the vessel orientation tensor estimated for each of the plurality of voxels. A centerline of the target vessel is extracted based on the estimated flow field for the plurality of vessels associated with the target vessel in the medical image by detecting a path that carries maximum flow.

Abstract: An image processing apparatus processes a first image signal including pixel signals, the first image signal having a picture portion in which a subject is shown and having a blank portion surrounding the picture portion. The apparatus includes: a blank detection unit configured to set similar frames different in size from one another within a display area of a screen representing the first image signal, in a stepwise manner from an outer periphery, and to detect the blank portion based on whether or not a part of the pixel signals belonging to a region outside each of the similar frames has the uniform color or brightness; a region-of-interest setting unit configured to set an initial region-of-interest corresponding to the picture portion in the first image signal, based on the blank portion; and a region-of-interest image generation unit configured to generate a region-of-interest image signal representing the initial region-of-interest.

Abstract: There is disclosed a system and method of performing facial recognition from RGB image data. The method includes generating a lower-resolution image from the RGB image data, performing a convolution of the lower-resolution image data to derive a probability map identifying probable facial regions and a probable non-facial regions, and performing a first deconvolution on the lower-resolution image using a bilinear interpolation layer to derive a set of coarse facial segments. The method further includes performing a second deconvolution on the lower-resolution image using a series of unpooling, deconvolution, and rectification layers to derive a set of fine facial segments, concatenating the set of coarse facial segments to the set of fine facial segments to create an image matrix made up of a set of facial segments, and generating a binary facial mask identifying probable facial regions and probable non-facial regions from the image matrix.

Abstract: A method of measuring animal health, such as rodent health, such as a degree of arthritis, is described. Steps include capturing video frames, then using optical flow to compute vector fields for each frame, then taking a maximum vector value over a first time period, then selecting a set of maximum vector values over a second time period, then clipping those values to a maximum, then computing an average of those values. This is an animal health metric. Metrics may be organized into a plot over time and may then be compared against known plots to compute an overall animal health and measure efficacy of treatments or other animal characteristics or behavior, including predictive.

Abstract: A terminal device includes: an acquiring unit that acquires image information; a displaying unit that displays the image information acquired by the acquiring unit; a receiving unit that receives an analysis position of image-quality unevenness from the image information displayed by the displaying unit; a detecting unit that detects pitch information by performing frequency analysis of the image information received by the receiving unit; a controlling unit that controls the displaying unit to display the pitch information detected by the detecting unit; and a storing unit that stores, as history information, the image information and the pitch information in association with the analysis position. The controlling unit controls the displaying unit to display the analysis position stored in the storing unit.

Abstract: A method and system for automatically inferring a subject's body position in a two-dimensional image produced by a medical-imaging system are disclosed. The image is labeled with a body position selected from a semantically meaningful set of candidate positions sequenced in order of their relative locations in a subject's body. A processor performs procedures that each identify a class of image features related to pixel intensity, such as a histogram of gradients, local binary patterns, or Haar-like features. A second set of procedures employs applications of a pretrained convolutional neural network that has learned to recognize features of a specific class of medical images. The results of both types of procedures are then mapped by a pretrained support-vector machine onto candidate image labels, which are mathematically combined into a single, semantically meaningful, label most likely to identify a body position of the subject shown by the image.

Abstract: An apparatus for producing a fundus image includes: a processor and a memory; an illumination component including a light source and operatively coupled to the processor; a camera including a lens and operatively coupled to the processor, wherein the memory stores instructions that, when executed by the processor, cause the apparatus to: execute an automated script for capture of the fundus image; and allow for manual capture of the fundus image.

Abstract: The present invention provides a cell evaluation device, a cell evaluation method and a non-transitory computer readable recording medium storing a cell evaluation program capable of evaluating an evaluation target cell. The cell evaluation device includes an image acquisition unit that acquires a cell image obtained by imaging a cell group; a cell evaluation unit that specifies an evaluation target cell and peripheral cells around the evaluation target cell in the cell group, and evaluates the evaluation target cell based on evaluation results of the peripheral cells; and a boundary setting unit that sets a boundary in the cell image based on a state of the cell group, in which when specifying the peripheral cells, the cell evaluation unit specifies only cells that are present in a divided region where the evaluation target cell is present among plural divided regions divided by the boundary, as the peripheral cells.

Abstract: A frequency decomposition unit performs frequency decomposition for a first radiographic image to acquire a plurality of band images. A scattered radiation removal unit calculates a scattered radiation component that is caused by a subject and is included in a second radiographic image which has a linear relationship with the amount of radiation, using the second radiographic image, and removes the scattered radiation component from the second radiographic image to acquire a scattered-radiation-removed radiographic image. A conversion function determination unit determines a conversion function for converting at least one of the band images according to the scattered radiation component included in the second radiographic image. A reconstruction unit converts the band image, using the conversion function, and generates a converted band image. The reconstruction unit reconstructs the scattered-radiation-removed radiographic image and the converted band image to acquire a processed radiographic image.

Abstract: A method for generating fluid flow images of a region of interest is disclosed. The method includes: scanning the region of interest to acquire a set of 4D-Flow magnetic resonance images, wherein the set comprises an anatomical magnitude image, a first velocity component image, a second velocity component image, and a third velocity component image; isolating the anatomical magnitude image, the first velocity component image, the second velocity component image, and the third velocity component image from the set of 4D-Flow magnetic resonance images; converting the first, second, and third velocity component images into a velocity vector field; modeling a location of an anatomical wall within the region of interest; calculating at least one flow dynamics parameter for the region of interest; and generating a visual representation of the anatomical wall and the at least one flow dynamics parameter.

Abstract: A method for the reconstruction of planning images is based on a plurality of first respiratory cycles of a patient and tomographic raw data recorded at the same time as the first respiratory cycles being received. A reference cycle is determined based on respiratory cycles of the patient, in particular based on the first respiratory cycles of the patient. A reference cycle is subdivided into temporally equidistant reference phases, a reference amplitude being associated with each reference phase. The reference amplitudes are not equidistantly scanned therefore. Furthermore, the first respiratory cycles of the patient are subdivided into first phases, the temporal positions of the first phases each being based on one of the reference amplitudes. Reconstruction of planning images based on first intervals of the raw data follows, wherein the first intervals correspond to the first phases.

Abstract: An eye detection method for detecting a degree of eye opening and closure of a testee includes detecting a location of an eye pupil, a location of an inner eye corner and a location of an outer eye corner of an eye of the testee; calculating an eye width according to the location of the inner eye corner and the location of the outer eye corner; multiplying the eye width by a specific ratio to obtain an eye height; generating a region of interest with a center on the location of the eye pupil, a length equal to the eye width, and a width equal to the eye height; and determining a ratio of the eye occupying a detection area in the region of interest, and determining the degree of eye opening and closure of the testee accordingly.

Abstract: A method of segmenting images of biological specimens using adaptive classification to segment a biological specimen into different types of tissue regions. The segmentation is performed by, first, extracting features from the neighborhood of a grid of points (GPs) sampled on the whole-slide (WS) image and classifying them into different tissue types. Secondly, an adaptive classification procedure is performed where some or all of the GPs in a WS image are classified using a pre-built training database, and classification confidence scores for the GPs are generated. The classified GPs with high confidence scores are utilized to generate an adaptive training database, which is then used to re-classify the low confidence GPs. The motivation of the method is that the strong variation of tissue appearance makes the classification problem more challenging, while good classification results are obtained when the training and test data origin from the same slide.

Abstract: A method for correcting a CT scan image is provided. A target image including a target region to be corrected is extracted from an original CT scan image. A target projection range corresponding to the target region is obtained by orthographically projecting the target image, and target projection data within the target projection range is corrected. Noise of the corrected target projection data is adjusted to generate noise-adjusted target projection data to improve noise consistency between target region and the rest of the CT scan image. A corrected CT scan image is reconstructed based on the noise-adjusted target projection data.

Abstract: A method of improving diagnostic and functional imaging is provided by obtaining at least two input images of a subject, using a medical imager, where each input image includes a different contrast, generating a plurality of copies of the input images using non-local mean (NLM) filtering, using an appropriately programmed computer, where each input image copy of the subject includes different spatial characteristics, obtaining at least one reference image of the subject, using the medical imager, where the reference image includes imaging characteristics that are different form the input images of the subject, training a deep network model, using data augmentation on the appropriately programmed computer, to adaptively tune model parameters to approximate the reference image from an initial set of the input and reference images, with the goal of outputting an improved quality image of other sets of low SNR low resolution images, for analysis by a physician.

Abstract: A medical imaging apparatus includes a scanner configured to acquire projection data of an object in a plurality of directions, and a data processor configured to generate a volume of interest based on the projection data, generate a two-dimensional (2D) reprojection image by reprojecting the volume of interest in at least one direction, and extract feature information from the 2D reprojection image.

Abstract: In an approach to tracking markers in one or more images, one or more computer processors identify objects that exist in more than one image from a plurality of images. The one or more computer processors analyze the identified objects for one or more physical characteristics. The one or more computer processors assign a marker to at least one object of the identified objects on at least one image of the more than one image, wherein the marker is annotated based upon the one or more physical characteristics of the object of the identified objects. The one or more computer processors store the more than one images, analysis data, and marker data associated with the identified objects and one or more markers. The one or more computer processors manipulate the one or more images based on a change in the objects across the more than one image.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for processing a CT (Computed Tomography) image are provided. An example method includes accessing an original CT image that is reconstructed from a first set of raw data and includes windmill artifacts, generating a high-frequency image by processing the original CT image, generating a low-frequency image by processing a plurality of thick images reconstructed from a second set of raw data and combining the plurality of processed thick images, the second set of raw data including the first set of raw data and each of the plurality of thick images including substantially no windmill artifacts, generating an intermediate image by synthesizing the high-frequency image and the low-frequency image, and obtaining a target CT image based on the generated intermediate image.

Abstract: Various systems and methods for implementing skin texture-based authentication are described herein. A system comprises a capture module to obtain at a wearable device worn by a user, an input representation of the user's skin; an analysis module to identify a set of features in the input representation; and an authentication module to authenticate the user based on the set of features.

Abstract: Intra-oral imaging apparatus and/or method embodiments can provide digital dental reflectance images used to identify selected areas of interest (AOI) within a dental region of interest (ROI) having selected image characteristics for both hemoglobin's total concentration and oxygenation level. In one embodiment, reflectance images can determine relative total hemoglobin using at least a wavelength band that includes an isobestic wavelength for absorption coefficients of oxyhemoglobin (HbO2) and deoxyhemoglobin (Hb).