Abstract: A data processing apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to output first complementary data by inputting, to a first neural network, first partial sampling data resulting from performing a partial sampling process; to obtain first corrected data, by performing a process to improve a consistency degree between the first complementary data and the first partial sampling data; to generate second partial sampling data, on the basis of the first corrected data and the first partial sampling data; and to output second complementary data, by inputting the second partial sampling data to a second neural network.

Abstract: A data processing device according to an embodiment includes a processing circuit. The processing circuit performs data processing using a learned model with a neural network including a division processing layer that divides input complex first vector data by complex second vector data containing features of the first vector data, a nonlinear layer that is disposed in the subsequent stage of the division processing layer and that performs a nonlinear operation, and a multiplication processing layer that is disposed in the subsequent stage of the nonlinear layer and that multiplies the input data by the second vector data.

Abstract: According to one embodiment, a reconstruction apparatus includes processing circuitry, the processing circuitry is configured to generate one or more sparse reconstructed images by performing image reconstruction on k-space data collected at each of a plurality of coils through under-sampling. The processing circuitry generates a first sensitivity map corresponding to each coil of the plurality of coils by using a first trained model. The processing circuitry performs a data consistency process for improving a degree of coincidence of data relating to the one or more sparse reconstructed images by using the first sensitivity map. The processing circuitry generates a full reconstructed image by performing a coil synthesis process using a second sensitivity map on the one or more sparse reconstructed images subjected to the data consistency process.

Abstract: A medical image processing apparatus according to the present embodiment includes processing circuitry. The processing circuitry inputs a first magnetic resonance image reconstructed with super-resolution processing on magnetic resonance data and a second magnetic resonance image obtained by imaging the same object as that of the first magnetic resonance image and with artifacts suppressed compared with the first magnetic resonance image, to a leaned model, the learned model being configured to output a third magnetic resonance image having the same resolution as that of the first magnetic resonance image and with the artifacts suppressed, generates the third magnetic resonance image based on the first magnetic resonance image and the second magnetic resonance image, using the learned model.

Abstract: An image generating apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to successively generate Magnetic Resonance (MR) images in a plurality of temporal phases by using a first reconstruction method. The processing circuitry is configured to determine a first temporal phase before MR images in all the temporal phases during a predetermined time period are generated. The processing circuitry is configured to generate an MR image in the first temporal phase determined by the determining unit, by using a second reconstruction method having a larger processing load than the first reconstruction method.

Abstract: A medical image processing apparatus according to one embodiment includes processing circuitry. The processing circuitry obtains input medical image data having a larger image region than output medical image data as a result of image processing; determines, in the input medical image data, a region that is smaller than the image region of the input medical image data and that is to affect image quality of the output medical image data when the image processing is applied to the input medical image data; performs the image processing to the determined region in the input medical image data; and outputs the output medical image data resulting from the image processing performed.

Abstract: A medical image processing apparatus according to one embodiment includes processing circuitry. The processing circuitry obtains input medical image data having a larger image region than that of output medical image data resulting from image processing; generates first intermediate image data by extracting from the input medical image data an image region including a region in the output medical image data and being smaller than the input medical image data; performs first image processing to the first intermediate image data, generates second intermediate image data based on the first intermediate image data having undergone the first image processing and on image data, of the input medical image data, containing an outside of an image region corresponding to the first intermediate image data; and converts the second intermediate image data to k-space data, and converts the k-space data having undergone k-spatial processing to the output medical image data.

Abstract: A medical image processing apparatus according to the present embodiment includes processing circuitry. The processing circuitry inputs a first magnetic resonance image reconstructed with super-resolution processing on magnetic resonance data and a second magnetic resonance image obtained by imaging the same object as that of the first magnetic resonance image and with artifacts suppressed compared with the first magnetic resonance image, to a leaned model, the learned model being configured to output a third magnetic resonance image having the same resolution as that of the first magnetic resonance image and with the artifacts suppressed, generates the third magnetic resonance image based on the first magnetic resonance image and the second magnetic resonance image, using the learned model.

Abstract: An MRI apparatus 1 includes sequence control circuitry 29 and processing circuitry 51. The sequence control circuitry 29 performs stack-of-stars data acquisition on an imaging region of a subject to acquire time-series k-space data. The processing circuitry 51 divides time-series k-space data into groups relating to a time direction, and calculates for each of the groups a motion feature amount of the imaging region based on k-space data of a k-space central portion. The processing circuitry 51 corrects for each of the groups the k-space data based on the motion feature amount and generates the corrected k-space data. The processing circuitry 51 reconstructs an MR image relating to the imaging region based on the corrected k-space data relating to the groups.

Abstract: An image generating apparatus according to the embodiment includes processing circuitry. The processing circuitry acquires MR data acquired in read-out directions including a first read-out direction and a second read-out direction intersecting the first read-out direction, filter sensitivity distributions corresponding to the read-out directions and indicating distributions of sensitivity of a low-pass filter, and coil sensitivity distributions corresponding to coil elements used to acquire the MR data. The processing circuitry generates synthesis sensitivity distributions for the respective read-out directions by synthesizing the filter sensitivity distributions and the coil sensitivity distributions for the respective read-out directions. The processing circuitry generates an MR image based on the synthesis sensitivity distributions and MR data.

Abstract: A medical image processing apparatus according to the present embodiment includes processing circuitry. The processing circuitry inputs a first magnetic resonance image reconstructed with super-resolution processing on magnetic resonance data and a second magnetic resonance image obtained by imaging the same object as that of the first magnetic resonance image and with artifacts suppressed compared with the first magnetic resonance image, to a leaned model, the learned model being configured to output a third magnetic resonance image having the same resolution as that of the first magnetic resonance image and with the artifacts suppressed, generates the third magnetic resonance image based on the first magnetic resonance image and the second magnetic resonance image, using the learned model.

Abstract: According to one embodiment, a medical data processing apparatus includes processing circuitry. The processing circuitry acquires medical data, and generates an imaging parameter by inputting the medical data to a trained model, the imaging parameter being a parameter of a medical image diagnostic apparatus with respect to the medical data, the trained model being trained to generate an imaging parameter of the medical image diagnostic apparatus based on medical data.

Abstract: A data processing device according to an embodiment includes a processing circuit. The processing circuit performs data processing using a learned model with a neural network including a division processing layer that divides input complex first vector data by complex second vector data containing features of the first vector data, a nonlinear layer that is disposed in the subsequent stage of the division processing layer and that performs a nonlinear operation, and a multiplication processing layer that is disposed in the subsequent stage of the nonlinear layer and that multiplies the input data by the second vector data.

Abstract: A medical image processing apparatus according to the present embodiment includes processing circuitry. The processing circuitry inputs a first magnetic resonance image reconstructed with super-resolution processing on magnetic resonance data and a second magnetic resonance image obtained by imaging the same object as that of the first magnetic resonance image and with artifacts suppressed compared with the first magnetic resonance image, to a leaned model, the learned model being configured to output a third magnetic resonance image having the same resolution as that of the first magnetic resonance image and with the artifacts suppressed, generates the third magnetic resonance image based on the first magnetic resonance image and the second magnetic resonance image, using the learned model.

Abstract: An image generating apparatus according to an embodiment includes processing circuitry. The processing circuitry obtains a coil sensitivity distribution indicating a sensitivity distribution of a reception coil used for an imaging process performed on an examined subject and magnetic resonance data acquired from the imaging process that is non-Cartesian and performed in a k-space. The processing circuitry performs, on the basis of the coil sensitivity distribution, registration between the coil sensitivity distribution and a gridding sensitivity distribution indicating a distribution of gridding sensitivity related to arranging the magnetic resonance data in the k-space. The processing circuitry generates a magnetic resonance image on the basis of a result of the registration, magnetic resonance data, coil sensitivity distribution, and gridding sensitivity distribution.

Abstract: A medical data processing device according to an embodiment includes a processing circuit. The processing circuit applies a linear calculation of a complex number coefficient and a non-linear activation not dependent on an argument of complex numbers, to medical data having a complex value.

Abstract: A data processing device according to an embodiment includes a processing circuit. The processing circuit includes a complex number neural network with an activation function by which an output varies according to an argument of complex.

Abstract: An image generating apparatus according to the embodiment includes processing circuitry. The processing circuitry acquires MR data acquired in read-out directions including a first read-out direction and a second read-out direction intersecting the first read-out direction, filter sensitivity distributions corresponding to the read-out directions and indicating distributions of sensitivity of a low-pass filter, and coil sensitivity distributions corresponding to coil elements used to acquire the MR data. The processing circuitry generates synthesis sensitivity distributions for the respective read-out directions by synthesizing the filter sensitivity distributions and the coil sensitivity distributions for the respective read-out directions. The processing circuitry generates an MR image based on the synthesis sensitivity distributions and MR data.

Abstract: An MRI apparatus according to the present embodiment acquires a plurality of MR signals corresponding to read-out directions including a first read-out direction and a second read-out direction intersecting the first read-out direction, and. The MRI apparatus includes processing circuitry and a low-pass filter. The processing circuitry specifies a signal area relating to generation of the MR signals in a subject. The processing circuitry sets a cutoff frequency defining a passband for the MR signals based on the signal area. The low-pass filter filters the MR signals acquired by scanning performed on the subject in the read-out directions using the cutoff frequency. The processing circuitry generates an MR image based on MR data generated by A/D conversion performed on the MR signals output from the low-pass filter.

Abstract: A magnetic resonance imaging apparatus according to an embodiment includes processing circuitry. The processing circuitry is configured to set a rotation angle of a non-Cartesian trajectory in a k-space on the basis of information related to cyclic movements of a subject to be imaged, to obtain k-space data by rotating the non-Cartesian trajectory at the set rotation angle, and to generate an image by reconstructing the k-space data.