The present disclosure is directed to MRI techniques. The techniques include occupying a central region of a first k-space with full sampling along a Cartesian trajectory, occupying a peripheral region of the first k-space with undersampling along a non-Cartesian trajectory; acquiring sensitivity distribution information of receiving coils; based on a sensitivity distribution chart, merging the Cartesian data of the central region according to multiple channels to obtain a third k-space; based on the sensitivity distribution chart, applying parallel imaging and compressed sensing to the undersampled non-Cartesian trajectory to reconstruct an image, obtaining a second k-space by transformation, and when the second k-space and third k-space are synthesized, using a central region of the second k-space to replace the third k-space of a corresponding region to obtain a k-space suitable for image reconstruction.

A method may include obtaining first image data relating to a region of interest (ROI) of a first subject. The first image data corresponding to a first equivalent dose level may be acquired by a first device. The method may also include obtaining a model for denoising relating to the first image data and determining second image data corresponding to an equivalent dose level higher than the first equivalent dose level based on the first image data and the model for denoising. In some embodiments, the method may further include determining information relating to the ROI of the first subject based on the second image data and recording the information relating to the ROI of the first subject.

A method for performing real-time magnetic resonance (MR) imaging on a subject is disclosed. A prep pulse sequence is applied to the subject to obtain a high-quality special subspace, and a direct linear mapping from k-space training data to subspace coordinates. A live pulse sequence is then applied to the subject. During the live pulse sequence, real-time images are constructed using a fast matrix multiplication procedure on a single instance of the k-space training readout (e.g., a single k-space line or trajectory), which can be acquired at a high temporal rate.

Improved (e.g., high-throughput, low-noise, and/or low-artifact) X-ray Microscopy images are achieved using a deep neural network trained via an accessible workflow. The workflow involves selection of a desired improvement factor (x), which is used to automatically partition supplied data into two or more subsets for neural network training. The neural network is trained by generating reconstructed volumes for each of the subsets. The neural network can be trained to take projection images or reconstructed volumes as input and output improved projection images or improved reconstructed volumes as output, respectively. Once trained, the neural network can be applied to the training data and/or subsequent data—optionally collected at a higher throughput—to ultimately achieve improved de-noising and/or other artifact reduction in the reconstructed volume.

Various methods and systems are provided for stationary CT imaging. In one embodiment, a method for an imaging system includes activating a plurality of emitters of a stationary distributed x-ray source unit to emit x-ray beams toward an object within an imaging volume, where the x-ray source unit does not rotate around the imaging volume, receiving attenuated x-ray beams with one or more detector arrays to form a sparse view projection dataset, where each attenuated x-ray beam generates a different view, and reconstructing an image from the sparse view projection dataset using a sparse view reconstruction method.

A system may transform sensor data from a sensor domain to an image domain using data-driven manifold learning techniques which may, for example, be implemented using neural networks. The sensor data may be generated by an image sensor, which may be part of an imaging system. Fully connected layers of a neural network in the system may be applied to the sensor data to apply an activation function to the sensor data. The activation function may be a hyperbolic tangent activation function. Convolutional layers may then be applied that convolve the output of the fully connected layers for high level feature extraction. An output layer may be applied to the output of the convolutional layers to deconvolve the output and produce image data in the image domain.

A tomographic imaging system, including a data processing component having a memory and at least one processor configured to: access scattering parameter data representing electromagnetic waves scattered by features within an object and originating from a plurality of antennas disposed around the object on a boundary S; process the scattering parameter data to generate a reconstructed image representing a spatial distribution of features within the object, said processing including: solving an electromagnetic inverse problem, wherein forward and inverse steps of the inverse problem are represented and solved as respective differential equations involving an electric field to determine values for the electric field; and process the determined values of the electric field to generate reconstructed image data representing one or more spatial distributions of one or more electromagnetic properties within the object.

A method and apparatus is provided that uses a deep learning (DL) network together with a multi-resolution detector to perform X-ray projection imaging to provide improved resolution similar to a single-resolution detector but at lower cost and less demand on the communication bandwidth between the rotating and stationary parts of an X-ray gantry. The DL network is trained using a training dataset that includes input data and target data. The input data includes projection data acquired using a multi-resolution detector, and the target data includes projection data acquired using a single-resolution, high-resolution detector. Thus, the DL network is trained to improve the resolution of projection data acquired using a multi-resolution detector. Further, the DL network is can be trained to additional correct other aspects of the projection data (e.g., noise and artifacts).

The present disclosure relates to systems, devices, and methods to reconstruct a three-dimensional model of an anatomy using two-dimensional images.

A method, system, and computer readable medium to compensate for consecutive missing views in Computed Tomography (CT) reconstruction. By utilizing at least one complementary ray from a previous or subsequent view, the missing view(s) can be filled in. When plural complementary rays exist, a linear or non-linear combination of rays can be used to fill in the missing views, and the weights used in the combination may be smoothed to prevent over-emphasis of the replacement views.