A method for acquiring image data obtains, for an intraoral feature, optical coherence tomography (OCT) data in three dimensions wherein at least one dimension is pseudo-randomly or randomly sampled and reconstructs an image volume of the intraoral feature using compressive sensing, wherein data density of the reconstructed image volume is larger than that of the obtained OCT data in the at least one dimension or according to a corresponding transform. The method renders the reconstructed image volume for display.

The present invention relates to a method for the evaluation of tissue gadolinium deposition that offers advantages compared with known methods. Comparison of different gadolinium-based contrast agents (GBCAs) based on retention, organ distribution, washout and safety is facilitated using the methods of the present invention.

Disclosed is a quantitative susceptibility mapping image processing method using an unsupervised learning-based neural network and an apparatus therefor. The quantitative susceptibility mapping image processing method includes receiving a phase image and a magnitude image for reconstructing the quantitative susceptibility mapping image, and reconstructing the quantitative susceptibility mapping image corresponding to the received phase image and the received magnitude image using an unsupervised learning-based neural network, and the neural network may be generated based on an optimal transport theory.

A method of imaging reconstruction includes providing a detector and a collimator, operating the detector to acquire a measured image of a target object from photons passing through the collimator, partitioning the collimator such that the collimator can be represented by a first matrix, providing an initial estimated image of the target object, and calculating an estimated image of the target object based on the measured image and the first matrix. The calculating of the estimated image includes an iteration using the initial estimated image as a starting point. The method also includes partitioning the collimator such that the collimator can be represented by a second matrix larger than the first matrix, and calculating a refined estimated image of the target object based on the measured image and the second matrix. The calculating of the refined estimated image includes an iteration using the estimated image as a starting point.

The present disclosure provides a system for image reconstruction. The system may obtain an initial image of a subject. The initial image may be generated based on scan data of the subject that is collected by an imaging device. The system may also generate a gradient image associated with the initial image. The system may further generate a target image of the subject by applying an image reconstruction model based on the initial image and the gradient image. The target image may have a higher image quality than the initial image.

A computer-implemented method for reconstruction of medical image data includes receiving medical measuring data, and minimizing a cost value via gradient descent. Minimizing the cost value includes: reconstructing the medical image data by applying a reconstruction function to the received medical measuring data in accordance with reconstruction parameters; determining a cost value by applying a cost function to the reconstructed medical image data; determining a gradient of the cost function with respect to the reconstruction parameters; adjusting the reconstruction parameters based on the gradient of the cost function with respect to the reconstruction parameters and the previous reconstruction parameters; and providing the adjusted reconstruction parameters. The acts of the minimizing are repeated until a termination condition is met. The reconstructed medical image data is provided.

Images are reconstructed from data acquired using an ultra-fast-pitch acquisition with a CT system. As an example, an ultra-fast-pitch acquisition mode in single-source helical CT (p≥1.5) can be used to acquire data. A trained machine learning algorithm, such as a neural network, is used to reconstruct images in which artifacts associated with insufficient data acquired in the ultra-fast-pitch mode are reduced. An example neural network can include customized functional modules using both local and non-local operators, as well as the z-coordinate of each image, to effectively suppress the location- and structure-dependent artifacts induced by the ultra-fast-pitch mode. The machine learning algorithm can be trained using a customized loss function that involves image-gradient-correlation loss and feature reconstruction loss.

A system and method include training of an artificial neural network to generate an output data set, the training based on the plurality of sets of emission data acquired using a first imaging modality and respective ones of data sets acquired using a second imaging modality.

The invention provides a multi-slice magnetic resonance imaging method and device based on long-distance attention model reconstruction. The method includes that: a deep learning reconstruction model is constructed; data preprocessing is performed on multiple slices of simultaneously acquired signals, and multiple slices of magnetic resonance images or K-space data is used as data input; learnable positional embedding and imaging parameter embedding are acquired; the preprocessed input data, the positional embedding and the imaging parameter embedding are input into the deep learning reconstruction model; and the deep learning reconstruction model outputs a result of the magnetic resonance reconstruction image. The invention further provides a device for implementing the method. The invention may improve the quality of the magnetic resonance image, improve the diagnosis accuracy of a doctor, increase the imaging speed, and improve the utilization rate of a magnetic resonance machine.

Methods and systems for generation and use of an accelerated tomographic reconstruction preconditioner (ATRP) for accelerated iterative tomographic reconstruction are disclosed. An example method for generating an ATRP for accelerated iterative tomographic reconstruction includes accessing data for a tomography investigation of a sample and determining a trajectory of the tomography investigation of a sample. At least one toy model sample depicting a feature characteristic of the sample are accessed and at least one candidate preconditioner is selected. A first performance of each of the at least one candidate preconditioner on the one or more toy samples is determined, where the candidate preconditioners are then updated to create updated candidate preconditioners. A second performance of each of the updated candidate preconditioners on the one or more toy samples is determined determining. An ATRP is then generated based on at least the first performance and the second performance.