Training procedure for artificial intelligence for mammographic data analysis for breast cancer detection including acquiring mammographic data from a plurality of sources and including mammographic images, report texts relating to images, and structured data obtained from SIO, EMR, BI-RADS and MOM including at least metadata relating to part of the images, processing the mammographic data through algorithms implementing a multimodal deep neural network (DNN) developing a mammographic data analysis model by performing learning based on sub-phases of first multi-label classification of each image implemented through a model with Encoder-Decoder architecture based on convolutional neural network (CNN) and/or Transformers, association of parts of report texts with images and/or parts of structured data, implemented through a model with Encoder-Decoder architecture based on a bidirectional long-term memory (Bi-LSTM) and/or Transformers, second multi-label classification of mammographic structured data implemented through a model with Encoder-Decoder architecture based on CNN and/or Transformers.

A method and system for acquiring, processing, storing, and displaying x-ray mammograms Mp tomosynthesis images Tr representative of breast slices, and x-ray tomosynthesis projection images Tp taken at different angles to a breast, where the Tr images are reconstructed from Tp images.

A method of generating a template image includes: receiving an input from a user representing identifications of an object in different respective slices of a volumetric image; using the input to determine a volume-of-interest (VOI) that includes voxels in a subset of the volumetric image; and determining the template image using at least some of the voxels in the VOI, wherein the act of determining the template image comprises performing a forward projection of the at least some of the voxels in the VOI using a processor. An image processing method includes: obtaining a volumetric image; performing forward projection of voxels in the volumetric image from different positions onto a first plane using a processor; and summing projections on the first plane resulted from the forward projection from the different positions to create a first image slice in the first plane.

Digital Tomosynthesis (DT) is a type of limited angle tomography providing the benefits of 3D imaging. Much like Computerized Tomography (CT), DT allows greater detection of 3D structures by viewing one slice at a time. In contrast to CT, the DT projection dataset is incomplete, which violates the tomographic sufficiency conditions and results in limited angle artefacts in the reconstructed images. The present invention provides a method of producing a tomogram in which reconstruction is performed along lines on the x-ray detector panel 20 defined by a point on the detector panel 20 closest to a point location of the x-ray emitter 10, to a location of a respective pixel on the perimeter of the x-ray detector panel 20. In this way, artefact reduction is achieved, particularly at higher cone beam angles, and at lower stand-off distances.

A method and apparatus is disclosed, which improves the analysis of an object within a scanned bag. Specifically, the techniques disclosed herein overcome the problem of measurement errors due to imaging artifacts, which can occur during imaging examinations like CT scans. This process also discloses a method of using an improved accuracy of data units of an object lead to more accurate classification of the material that makes up the object.

A method for immersively displaying a scanned environment of a region to a set of users in a training environment wearing augmented reality head display units. The training environment includes a pseudo-GPS system, which allows position tracking over time. This enables rehearsing military operations before they occur.

An approach for improving determining a significant slice associated with a tumor from a volume of medical images is disclosed. The approach is based on the annotation of tumor range and the slice index in which the tumor appears to have the largest area. The approach infer a tumor growth classifier on sliding window of the volume slices and creates a discrete integral function out of the classifier predictions. The approach applies post processing on the discrete integral function which can include a smoothing function and a bias correction. The approach selects the slice index of maximum value from the post processing step.

An object detection method is suitable for a 3D mammogram. The object detection method comprises steps of: controlling N filers to execute a filtering computation in the 3D mammogram respectively to generate N 3D filtering images; computing a difference variation among the plurality of voxels to obtain a blurriness value of the plurality of voxels; using the blurriness value of the plurality of voxels in a decision module to execute a plurality of first decision operators to generate a plurality of first decision results, and using one of the plurality of first decision results to execute the plurality of second decision operators to generate a plurality of second decision results; and executing a final decision operator by using the plurality of first decision results and the plurality of second decision results to generate a detection object of the 3D mammogram.

A composite image (300) is created from a plurality of tomographic slices by creating a plurality of two-dimensional slice images by projecting the slices, dividing each slice image into tiles (302) according to a pattern (304), calculating a focus value for each tile (302) of each slice image, selecting one tile (302) for each position in the pattern having a highest focus measure value, and assembling the composite image (300) from said selected tiles (302).

An image display apparatus includes a depth map creating unit that creates, on the basis of a two-dimensional radiation image and a plurality of tomographic images for the same subject, a plurality of depth maps in which each position on the two-dimensional radiation image and depth information indicating a depth directional position of a tomographic plane corresponding to each position are associated with each other while changing a correspondence relationship between each position on the two-dimensional radiation image and the depth information.