Motion estimation and correction in magnetic resonance imaging

1. A medical system comprising:

a memory storing machine executable instructions, contains an intermediate magnetic resonance image reconstruction neural network and a spatial transformation estimation module, wherein the spatial transformation estimation module is configured for outputting spatial transform data descriptive of a spatial transform between a reference k-space data group and multiple k-space data groups in response to receiving the reference k-space data group and the multiple k-space data groups as input; wherein the intermediate magnetic resonance imaging reconstruction neural network is configured to output an intermediate magnetic resonance image in response to receiving a respective k-space data group; and

a computational system, wherein execution of the machine executable instructions causes the computational system to:

receive k-space data acquired according to a magnetic resonance imaging protocol by a Cartesian sampling pattern, wherein the k-space data is divided into the multiple k-space data groups that are disjoint in k-space;

select one of the multiple k-space data groups as the reference k-space data group;

reconstruct respective intermediate magnetic resonance images for the multiple k-space data groups and the reference k-space data group by inputting the multiple k-space data groups and the reference k-space data group into the intermediate magnetic resonance imaging reconstruction neural network;

calculate the spatial transform data for each of the multiple k-space data groups by inputting the respective intermediate magnetic resonance images for the multiple k-space data groups and the reference k-space data group into the spatial transformation estimation module for making a registration in image space between the respective intermediate magnetic resonance images for the multiple k-space data groups and the respective intermediate magnetic resonance image for the reference k-space data group; and

arrange to reconstruct a corrected magnetic resonance image according to the magnetic resonance imaging protocol using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

2. The medical system of claim 1 , wherein the reconstruction of the corrected magnetic resonance image is at least partially performed by calculating corrected multiple k-space data groups using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

3. The medical system of claim 1 , wherein the reconstruction of the corrected magnetic resonance image is performed as an optimization problem.

4. The medical system of claim 1 , wherein the memory further contains a final magnetic resonance image reconstruction neural network configured to output a corrected magnetic resonance image in response to receiving multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups, wherein the reconstruction of the corrected magnetic resonance image is performed by inputting the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups into the final magnetic resonance image reconstruction neural network.

5. The medical system of claim 1 , wherein execution of the machine executable instructions further causes the computational system to divide the k-space data into the multiple k-space data groups according to any one of the following: by acquisition shots, by an acquisition time, by acquisition channel, a motion signal, and combinations thereof.

6. The medical system of claim 1 , wherein the registration comprises a rigid body registration.

7. The medical system of claim 1 , wherein the magnetic resonance imaging protocol is a parallel imaging magnetic resonance imaging protocol.

8. The medical system of claim 1 , wherein the medical system further comprises a magnetic resonance imaging system, wherein the memory further contains pulse sequence commands configured for acquiring the k-space data according to the magnetic resonance imaging protocol, wherein execution of the machine executable instructions further causes the computational system to acquire the k-space data by controlling the magnetic resonance imaging system with the pulse sequence commands.

9. The medical system of claim 8 , wherein the magnetic resonance imaging system comprises a radio frequency system with multiple receive channels, wherein the magnetic resonance imaging protocol is a parallel imaging magnetic resonance imaging protocol, wherein execution of the machine executable instructions further causes the computational system to divide the k-space data into the multiple k-space data groups at least partially according to the receive channels.

10. The medical system of claim 8 , wherein the machine executable instructions further causes the computational system to divide the k-space data into the multiple k-space data groups at least partially according to any one of the following:

by acquisition shots;

by an acquisition time or acquisition interval; and

combinations thereof.

11. The medical system of claim 8 , wherein the medical system further comprises a subject motion monitor configured for generating subject motion data descriptive of a motion or position of the subject, wherein execution of the machine executable instructions further causes the computational system to:

control the subject motion monitor to acquire the subject motion data during acquisition of the k-space data; and

divide the k-space data into the multiple k-space data groups at least partially according to the subject motion data.

12. The medical system of claim 1 , wherein the spatial transformation estimation module comprises a spatial transformation estimating neural network.

13. A method of medical imaging, wherein the method comprises:

receiving k-space data acquired by a Cartesian sampling pattern according to a magnetic resonance imaging protocol, wherein the k-space data is divided into multiple k-space data groups that are disjoint in k-space;

selecting one of the multiple k-space data groups as a reference k-space data group;

reconstructing respective intermediate magnetic resonance images for the multiple k-space data groups and the reference k-space data group by inputting the multiple k-space data groups and the reference k-space data group into an intermediate magnetic resonance imaging reconstruction neural network;

calculating spatial transform data for each of the multiple k-space data groups using the respective intermediate magnetic resonance images by making a registration in image space between the respective intermediate magnetic resonance images for the multiple k-space data groups and the respective intermediate magnetic resonance image for the reference k-space data group; and

reconstructing a corrected magnetic resonance image according to the magnetic resonance imaging protocol using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

14. The method of claim 13 , wherein reconstructing the corrected magnetic resonance image comprises calculating corrected multiple k-space data groups using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

15. The method of claim 13 , wherein reconstructing the corrected magnetic resonance image comprises performing an optimization problem.

16. The method of claim 13 , further comprising dividing the k-space data into the multiple k-space data groups according to at least one of acquisition shots, an acquisition time, acquisition channel, or a motion signal.

17. The method of claim 13 , wherein the registration comprises rigid body registration.

18. The method of claim 13 , wherein the magnetic resonance imaging protocol is a parallel imaging magnetic resonance imaging protocol.

19. A non-transitory computer readable medium storing machine executable instructions for execution by a computational system, wherein execution of the machine executable instructions causes the computational system to:

receive k-space data acquired by a Cartesian sampling pattern according to a magnetic resonance imaging protocol, wherein the k-space data is divided into multiple k-space data groups that are disjoint in k-space;

select one of the multiple k-space data groups as a reference k-space data group;

reconstruct respective intermediate magnetic resonance images for the multiple k-space data groups and the reference k-space data group by inputting the multiple k-space data groups and the reference k-space data group into an intermediate magnetic resonance imaging reconstruction neural network;

calculate spatial transform data for each of the multiple k-space data groups using the respective intermediate magnetic resonance images by making a registration in image space between the respective intermediate magnetic resonance images for the multiple k-space data groups and the respective intermediate magnetic resonance image for the reference k-space data group; and

arrange to reconstruct a corrected magnetic resonance image according to the magnetic resonance imaging protocol using the multiple k-space data groups and the spatial transform data for each of the multiple k-space data groups.

20. The non-transitory computer readable medium of claim 19 , wherein execution of the machine executable instructions further causes the computational system to divide the k-space data into the multiple k-space data groups according to at least one of acquisition shots, an acquisition time, acquisition channel, or a motion signal.