A method and system are provided for detecting a position of a periodically moving organ in a MRI examination. MR images of an examining person including a periodically moving organ are provided over a plurality of periodic cycles of the periodically moving organ. Based on the provided MR images, a pixel frequency is associated with each pixel of the MR images. Using the associated pixel frequencies and the positions of the pixels within the MR images, the position and the frequency of the periodically moving organ are determined.

The invention provides for a magnetic resonance imaging system (100, 200) comprising a memory (148) for storing machine executable instructions (150) and pulse sequence commands (152). The pulse sequence commands are configured for acquiring a four dimensional magnetic resonance data set (162) from an imaging region of interest (109). The four dimensional magnetic resonance data set is at least divided into three dimensional data magnetic resonance data sets (400, 402, 404, 406, 408) indexed by a repetitive motion phase of the subject. The three dimensional data magnetic resonance data sets are further at least divided into and indexed by k-space portions (410, 412, 414, 416, 418, 420, 422, 424, 426, 428, 430, 432, 434, 436). The magnetic resonance imaging system further comprises a processor (144) for controlling the magnetic resonance imaging system. Execution of the machine executable instructions causes the processor during a first operational portion (310) to iteratively: receive (300) a motion signal (156) descriptive of the repetitive motion phase; acquire (302) an initial k-space portion using the pulse sequence commands, wherein the initial k-space portion is selected from the k-space portions; store (304) the motion signal and the initial k-space portion in a buffer (158) for each iteration of the first operational portion; at least partially construct (306) a motion phase mapping (160) between the motion signal and the repetitive motion phase; and continue (308) the first operational portion until the motion phase mapping is complete. Execution of the machine executable instructions causes the processor to assign (312) the initial k-space portion for each iteration of the first operational portion in the temporary buffer to the four dimensional magnetic resonance data set using the motion phase mapping. Execution of the machine executable instructions causes the processor during a second operational portion (332) to iteratively: receive (314) the motion signal; determine (316) a predicted next motion phase using the motion signal and the motion phase mapping; select (318) a subsequent k-space portion (154) from the k-space portions of the four dimensional magnetic resonance data set using the predicted next motion phase; acquire (320) the subsequent k-space portion using the pulse sequence commands; rereceive (322) the motion signal; determine (324) a current motion phase using the re-received motion signal and the motion phase mapping; assign (326) the

An image reconstruction method according to an embodiment includes: collecting first k-space data at a first time and second k-space data having an undersampled pattern different from an undersampled pattern of the first k-space data at a second time different from the first time; generating intermediate data by converting data including the first k-space data and the second k-space data; generating, by inversely converting the intermediate data, third k-space data and fourth k-space data that correspond to the first k-space data and the second k-space data, respectively, and in each of which at least part of a region undersampled through the corresponding undersampled pattern is filled; and generating a magnetic resonance image at a time different from any of the first time and the second time by converting k-space data obtained by combining at least part of the third k-space data and at least part of the fourth k-space data.

A trained deep learning network is for determining a cardiac phase in magnet resonance imaging. In an embodiment, the trained deep learning network includes an input layer; an output layer; and a number of hidden layers between input layer and output layer, the layers processing input data entered into the input layer. In an embodiment, the deep learning network is designed and trained to output a probability or some other label of a certain cardiac phase at a certain time from entered input data. A method for determining a cardiac phase in magnet resonance imaging; a related device; a training method for the deep learning network; a control device and a related magnetic resonance imaging system are also disclosed.

Apparatus and techniques are described herein for nuclear magnetic resonance (MR) projection imaging. Such projection imaging may be used for generating four-dimensional (4D) imaging information representative of a physiologic cycle of a subject, such as including generating two or more two-dimensional (2D) images, the 2D images comprising projection images representative of different projection angles, and the 2D images generated using imaging information obtained via nuclear magnetic resonance (MR) imaging, assigning the particular 2D images to bins at least in part using information indicative of temporal positions within the physiologic cycle corresponding to the particular 2D images, constructing three-dimensional (3D) images using the binned 2D images, and constructing the 4D imaging information, comprising aggregating the 3D images.

Systems and methods are provided for performing automated reconstruction of a dynamic MRI dataset that is acquired without a fixed temporal resolution. On one or more image quality metrics (IQMs) are obtained by processing a subset of the acquired dataset. In one example implementation, at each stage of an iterative process, one or more IQMs of the image subset is computed, and the parameters controlling the reconstruction and/or the strategy for data combination are adjusted to provide an improved or optimal image reconstruction. Once the IQM of the image subset satisfies acceptance criteria based on an estimate of the overall temporal fidelity of the reconstruction, the full reconstruction can be performed, and the estimate of the overall temporal fidelity can be reported based on the IQM at the final iteration.

A system and method is provided for acquiring flow encoded data from a subject using a magnetic resonance imaging (MRI) system. The method includes acquiring flow encoded (FE) data with alternating encoding polarities and along two of three orthogonal directions through the subject over at least two cycles of the flow within the subject; and separating the FE data into directional FE datasets using a temporal filter that separates the FE data based on temporal modulation of the FE directions caused by the alternating encoding polarities extending over the at least two cycles of the flow within the subject that shift the Fourier spectrum of velocity waveforms corresponding to the FE data. The method also includes using the directional FE datasets to generate an image of the subject showing flow within the subject caused by the at least two cycles of flow within the subject.

In various embodiments, the present invention teaches methods and related systems for imaging the coronary arteries in high spatiotemporal resolution for the assessment of coronary stenosis. In some embodiments, the method teaches the use of a 3D radial k-space trajectory, continuous acquisition, retrospective cardiac and respiratory self-gating, and non-rigid cardiac and respiratory motion correction to reconstruct any arbitrary cardiac phase with minimal motion artifacts and high image quality.

A method for determining time periods of minimal motion of a physiologic organ includes monitoring a physiologic triggering signal associated with a patient and using an MRI cine pulse sequence to acquire a temporal series of projections of the organ. The temporal series is analyzed to determine times relative to a physiologic triggering signal during which motion of the organ is below a threshold. Motion is assessed by first creating a signal intensity versus time curve of one pixel or an average of multiple pixels included in the temporal series. A noise filter and normalization is applied to the signal intensity versus time curve to yield a filtered and normalized time curve. The temporal derivative of the filtered and normalized time curve is determined. The absolute value of the motion-analog function is evaluated for being smaller than the threshold to determine the times where motion is below the threshold.

An imaging method may include obtaining imaging data associated with a region of interest (ROI) of an object. The imaging data may correspond to a plurality of time-series images of the ROI. The imaging method may also include determining, based on the imaging data, a data set including a spatial basis and one or more temporal bases. The spatial basis may include spatial information of the imaging data. The one or more temporal bases may include temporal information of the imaging data. The imaging method may also include storing, in a storage medium, the spatial basis and the one or more temporal bases.