Abstract: Systems and methods for image reconstruction for parallel MR imaging are disclosed that receive a k-space single-coil measurement dataset that includes at least two k-space single-coil measurement sets, transforming the k-space single-coil measurement dataset to an estimated CSM using a coil sensitivity estimation module, and transforming the k-space single-coil measurement dataset and the estimated CSM into a final MR image using an MRI reconstruction module. In some aspects, the coil sensitivity estimation module and MRI reconstruction module include deep learning neural networks trained without the use of ground truth data.

Abstract: A method for reconstructing dynamic contrast-enhanced (DCE) MR images includes receiving a plurality of continuous free-breathing DCE images, the plurality of images obtained with a contrast, sorting the images by identifying a respiratory phase associated with each of the continuous free-breathing DCE images, reconstructing the plurality continuous free-breathing DCE images into a 4D respiratory motion-resolved image, obtaining 3D deformable motion vector fields (MVFs), and utilizing a deep learning based motion transformation integrated forward-Fourier (DL-MOTIF) model and the 3D deformable MVFs to reconstruct the DCE MR images.

Abstract: A method for reconstructing dynamic contrast-enhanced (DCE) MR images includes receiving a plurality of continuous free-breathing DCE images, the plurality of images obtained with a contrast, sorting the images by identifying a respiratory phase associated with each of the continuous free-breathing DCE images, reconstructing the plurality continuous free-breathing DCE images into a 4D respiratory motion-resolved image, obtaining 3D deformable motion vector fields (MVFs), and utilizing a deep learning based motion transformation integrated forward-Fourier (DL-MOTIF) model and the 3D deformable MVFs to reconstruct the DCE MR images.

Abstract: A computer-implemented method of reconstructing magnetic resonance (MR) images of a subject is provided. The method includes executing a neural network model for analyzing MR images, wherein the neural network model is trained with a first subset of training MR images as inputs and a second subset of the training MR images as outputs, wherein each image in the first subset is acquired during a neighboring respiratory phase of at least one of the images in the second subset. The method further includes receiving MR signals, reconstructing crude MR images based on the MR signals, analyzing the crude MR images using the neural network model, deriving clear MR images based on the analysis, wherein the clear MR images include reduced artifacts, compared to the crude MR images, and outputting the clear MR images.

Abstract: Systems and methods for image reconstruction for parallel MR imaging are disclosed that receive a k-space single-coil measurement dataset that includes at least two k-space single-coil measurement sets, transforming the k-space single-coil measurement dataset to an estimated CSM using a coil sensitivity estimation module, and transforming the k-space single-coil measurement dataset and the estimated CSM into a final MR image using an MRI reconstruction module. In some aspects, the coil sensitivity estimation module and MRI reconstruction module include deep learning neural networks trained without the use of ground truth data.

Abstract: A computer-implemented method of reconstructing magnetic resonance (MR) images of a subject is provided. The method includes executing a neural network model for analyzing MR images, wherein the neural network model is trained with a first subset of training MR images as inputs and a second subset of the training MR images as outputs, wherein each image in the first subset is acquired during a neighboring respiratory phase of at least one of the images in the second subset. The method further includes receiving MR signals, reconstructing crude MR images based on the MR signals, analyzing the crude MR images using the neural network model, deriving clear MR images based on the analysis, wherein the clear MR images include reduced artifacts, compared to the crude MR images, and outputting the clear MR images.

Abstract: Systems and methods for producing motion-corrected MR images that include an MRI scanner with a plurality of coils in which the plurality of coils is configured to detect a plurality of MR imaging datasets are provided. Each MR imaging dataset includes a plurality of MR signals detected by one coil of the plurality of coils. The system further includes a controller operatively coupled to the MRI scanner that includes at least one processor and a non-volatile memory. The controller further includes an image processing unit configured to receive the plurality of MR imaging datasets associated with the plurality of coils, identify a motion cycle using an automated motion detection method, and to partition the plurality of MR imaging datasets into a plurality of image groups, in which each image group comprises a portion of the plurality of MR imaging datasets associated with one of the phases of the motion cycle.

Abstract: Systems and methods for producing pseudo-CT images using a dual flip angle multi-echo ultra-short echo time (DUFA-MUTE) MRI method are disclosed. The DUFA-MUTE MRI imaging method includes obtaining MR signals according to a DUFA-MUTE MRI sequence that includes first and second multiple ultrashort echo time (MUTE) sequence characterized by first and second flip angles FA1/FA2, and in which both MUTE sequences obtain MR signals at first and second echo times TE1/TE2. HU values are assigned to each imaged voxel based on each voxel's R1 value calculated from the MR signals, as well as each voxel's assigned tissue type. The imaged voxels and assigned HU values are combined to produce a pseudo-CT image. Pseudo-CT images optionally form the basis for attenuation maps suitable for use in combined PET/MRI systems and/or electron density maps suitable for use in radiation therapy systems.

Abstract: Systems and methods for producing pseudo-CT images using a dual flip angle multi-echo ultra-short echo time (DUFA-MUTE) MRI method are disclosed. The DUFA-MUTE MRI imaging method includes obtaining MR signals according to a DUFA-MUTE MRI sequence that includes first and second multiple ultrashort echo time (MUTE) sequence characterized by first and second flip angles FA1/FA2, and in which both MUTE sequences obtain MR signals at first and second echo times TE1/TE2. HU values are assigned to each imaged voxel based on each voxel's R1 value calculated from the MR signals, as well as each voxel's assigned tissue type. The imaged voxels and assigned HU values are combined to produce a pseudo-CT image. Pseudo-CT images optionally form the basis for attenuation maps suitable for use in combined PET/MRI systems and/or electron density maps suitable for use in radiation therapy systems.

Abstract: Systems and methods for producing motion-corrected MR images that include an MRI scanner with a plurality of coils in which the plurality of coils is configured to detect a plurality of MR imaging datasets are provided. Each MR imaging dataset includes a plurality of MR signals detected by one coil of the plurality of coils. The system further includes a controller operatively coupled to the MRI scanner that includes at least one processor and a non-volatile memory. The controller further includes an image processing unit configured to receive the plurality of MR imaging datasets associated with the plurality of coils, identify a motion cycle using an automated motion detection method, and to partition the plurality of MR imaging datasets into a plurality of image groups, in which each image group comprises a portion of the plurality of MR imaging datasets associated with one of the phases of the motion cycle.