A method for automated detection of landmarks from 3D medical image data using deep learning according to the present inventive concept, the method includes receiving a 3D volume medical image, generating a 2D intensity value projection image based on the 3D volume medical image, automatically detecting an initial anatomical landmark using a first convolutional neural network based on the 2D intensity value projection image, generating a 3D volume area of interest based on the initial anatomical landmark and automatically detecting a detailed anatomical landmark using a second convolutional neural network different from the first convolutional neural network based on the 3D volume area of interest.

Disclosed are an apparatus and method for medical image processing according to pathologic lesion properties, the method including: recognizing a readout area different from an original readout area in a medical image by applying a previously trained deep learning model to the medical image, extracting properties, which include at least one of a location and a size of the readout area, from the medical image, and generating a readout image for the readout area, which is different from the original readout area corresponding to a purpose of taking the medical image, by reconstructing the medical image, thereby having an effect on generating a readout image for a different kind of pathologic lesion from a previously acquired medical image.

[Problem] To perform interpretation assistance of a back portion image in real time using a general-purpose computer for any case of scoliosis.

[Solution] An interpretation assistance apparatus includes a center line creating unit that creates a center line C in a back portion image of a subject, a measurement interval designating unit that accepts designation of an arbitrary measurement interval I along the center line C, a measurement width designating unit that accepts designation of an arbitrary measurement width W toward intersecting directions of the center line C, and a measurement point coordinate obtaining unit that obtains depth coordinates of respective measurement points that are apart from the center line on both the right and left sides by the measurement width W at every measurement interval I.

An image processing method is provided. An image including a target object is obtained. Image segmentation is performed on the image. A mask image of the target object is determined based on the image segmentation performed on the image. A first feature extraction is performed on the image. A first predicted value associated with the target object is determined based on a first feature extraction result of the first feature extraction performed on the image. A second feature extraction is performed on the mask image. A second predicted value associated with the target object is determined based on a second feature extraction result of the second feature extraction performed on the mask image. A target predicted value associated with the target object is determined according to the first predicted value and the second predicted value.

An imaging unit of an MRI apparatus performs imaging of a positioning image of a subject including a spine; a first imaging that images a cross section including the spine and extending along a longitudinal direction of the spine; and a second imaging that images a cross section in a direction of traversing the spine. An automatic cross-section position setting unit detects a specific tissue of the spine using a scout image or an image including the spine acquired in the first imaging step, performs a matching process between the detected specific tissue of the spine and a spine model, and calculates an imaging cross-section position of the second imaging based upon a specific tissue position of the spine specified by matching, thereby performing automatic setting.

A method of modeling lungs of a patient includes acquiring computed tomography data of a patient's lungs, storing a software application within a memory associated with a computer, the computer having a processor configured to execute the software application, executing the software application to differentiate tissue located within the patient's lung using the acquired CT data, generate a 3-D model of the patient's lungs based on the acquired CT data and the differentiated tissue, apply a material property to each tissue of the differentiated tissue within the generated 3-D model, generate a mesh of the 3-D model of the patient's lungs, calculate a displacement of the patient's lungs in a collapsed state based on the material property applied to the differentiated tissue and the generated mesh of the generated 3-D model, and display a collapsed lung model of the patient's lungs based on the calculated displacement of the patient's lungs.

The present disclosure provides a computer-implemented method, a device, and a computer program product for radiographic bone mineral density (BMD) estimation. The method includes receiving a plain radiograph, detecting landmarks for a bone structure included in the plain radiograph, extracting an ROI from the plain radiograph based on the detected landmarks, estimating the BMD for the ROI extracted from the plain radiograph by using a deep neural network.

A vertebra localization and identification method includes: receiving one or more images of vertebrae of a spine; applying a machine learning model on the one or more images to generate three-dimensional (3-D) vertebra activation maps of detected vertebra centers; performing a spine rectification process on the 3-D vertebra activation maps to convert each 3-D vertebra activation map into a corresponding one-dimensional (1-D) vertebra activation signal; performing an anatomically-constrained optimization process on each 1-D vertebra activation signal to localize and identify each vertebra center in the one or more images; and outputting the one or more images, wherein on each of the one or more outputted images, a location and an identification of each vertebra center are specified.

Aspects of the present disclosure relate to systems, devices and methods for performing a surgical step or surgical procedure for example with visual guidance using a head mounted display or with a surgical navigation system or with a surgical robot. A computer processor can be configured to determine the pose of a first vertebra with an attached first marker and a second vertebra with an attached second marker. The computer processor can be configured to determine the pose of at least one vertebra interposed or adjacent to the first and second vertebrae with attached markers, e.g. fiducial markers.

The technology relates to generating a three-dimensional point cloud model of an anatomical structure. A computer accessible memory stores a three-dimensional array of data elements describing multiple anatomical features of a subject, each of the data elements having associated therewith positional data and a separate parameter value. A processor may be configured to identify any data elements in the three-dimensional array having an associated parameter value satisfying a predefined threshold value associated with at least one anatomical feature. The processor may be further configured to generate a visually displayable three-dimensional point cloud model of at least one anatomical structure having a first plurality of points in the point cloud model which define an exterior perimeter of the at least one anatomical structure and a second plurality points in the point cloud model which define at least one feature interior of the exterior perimeter of the at least one anatomical structure.