Methods, apparatus and systems for deep learning based image reconstruction are disclosed herein. An example at least one computer-readable storage medium includes instructions that, when executed, cause at least one processor to at least: obtain a plurality of two-dimensional (2D) tomosynthesis projection images of an organ by rotating an x-ray emitter to a plurality of orientations relative to the organ and emitting a first level of x-ray energization from the emitter for each projection image of the plurality of 2D tomosynthesis projection images; reconstruct a three-dimensional (3D) volume of the organ from the plurality of 2D tomosynthesis projection images; obtain an x-ray image of the organ with a second level of x-ray energization; generate a synthetic 2D image generation algorithm from the reconstructed 3D volume based on a similarity metric between the synthetic 2D image and the x-ray image; and deploy a model instantiating the synthetic 2D image generation algorithm.

Disclosed herein is a system and method, which utilize primary beam modulation to enable single-scan dual-energy CT (DECT) on a conventional CT scanner. An attenuation sheet with a spatially-varying pattern is placed between the x-ray source and the imaged object. During the CT scan, the modulator selectively hardens the x-ray beam at specific detector locations. Thus, this method simultaneously acquires high and low energy data at each projection angle. High and low energy CT images can then reconstructed from the projections via an iterative CT reconstruction algorithm, which accounts for the spatial modulation of the projected x-rays.

In a method and system for reconstructing computed tomography image data in which CT image data is de-noised. Then simulated noise is added, followed by another de-noising step to estimate the bias. Then, the estimated bias information is used to correct the original de-noised image data to arrive at second pass image data.

A computer-implemented method for image processing is provided. The method includes acquiring multiple multi-energy spectral scan datasets and computing basis material images representative of multiple basis materials from the multi-energy spectral scan datasets, wherein the multiple basis material images include correlated noise. The method also includes jointly denoising the multiple basis material images in at least a spectral domain utilizing a deep learning-based denoising network to generate multiple de-noised basis material images.

A method comprising: receiving a radiographic image dataset representing a sequential radiographic scan of a region of a human subject; receiving three-dimensional (3D) image data representing an optical scan of a surface of said region, wherein said 3D image data is performed simultaneously with said sequential radiographic scan; estimating a time-dependent motion of said subject during said acquisition, relative to a specified position, based, at least in part, on said 3D image data; and using said estimating to determine corrections for said radiographic image dataset, based, at least in part, on a known transformation between corresponding coordinate systems of said radiographic image dataset and said 3D image data.

Extended field-of-view imaging is enabled by combined imaging with a kilovolt (“kV”) x-ray source and a megavolt (“MV”) x-ray source, in which at least one of the corresponding x-ray detectors is laterally offset from the target isocenter by an amount such that the x-ray detector does not have a view of the target isocenter. This scan geometry enables the reconstruction of non-truncated images without resorting to the more expensive solution of outfitting the imaging or radiotherapy system with enlarged x-ray detectors.

A method is disclosed for generating an image. An embodiment of the method includes detecting a first projection data set via a first group of detector units, the first group including a first plurality of first detector units, each having more than a given number of detector elements; detecting a second projection data set via a second group of detector units, the second group including a second plurality of second detector units, each including, at most, the given number of detector elements; reconstructing first image data based on the first projection data set; reconstructing second image data based on the second projection data set; and combining the first image data and the second image data. A non-transitory computer readable medium, a data processing unit, and an imaging device including the data processing unit are also disclosed.

There is provided a method and corresponding system for image reconstruction based on energy-resolved x-ray data. The method comprises collecting (S1) at least one representation of energy-resolved x-ray data, and performing (S2) at least two basis material decompositions based on said at least one representation of energy-resolved x-ray data to generate at least two original basis image representation sets. The method further comprises obtaining or selecting (S3) at least two basis image representations from at least two of said original basis image representation sets, and processing (S4) said obtained or selected basis image representations by data processing based on machine learning to generate at least one representation of output image data.

A method for generating attenuation map is disclosed. The method includes acquiring an anatomic image and PET data indicative of a subject, wherein the anatomic image comprises a plurality of voxels. The method also includes fetching a reference image to register the anatomic image, the reference image includes voxel segmentation information. The method further includes segmenting the anatomic image into a plurality of regions based on the voxel segmentation information. The method further includes generating a first attenuation map corresponding to the anatomic image by assigning attenuation coefficients to the plurality of regions. The method further includes calculating a registration accuracy between the anatomic image and the reference image. The method further includes determining a probability distribution of attenuation coefficient. The method further includes updating the first attenuation map iteratively based on the probability distribution of attenuation coefficient and the PET data to obtain a final attenuation map.

A method includes acquiring a first dataset of projection measurements at a first energy spectrum and a second dataset of projection measurements at a second energy spectrum different from the first energy spectrum by switching between acquiring the first dataset for a set number of consecutive views at different projection angles at the first energy spectrum and acquiring the second dataset for the set number of consecutive views at different projection angles at the second energy spectrum. The set number of consecutive views is greater than one. The method includes supplementing both the first dataset with estimated projection measurements at the first energy spectrum and the second dataset with estimated projection measurements at the second energy spectrum to provide missing projection measurements at different projection angles not acquired during the imaging scan for the first dataset and the second dataset.