Accelerating magnetic resonance imaging using parallel imaging and iterative image reconstruction
The present disclosure provides various systems and methods for magnetic resonance imaging. In one aspect, a method for magnetic resonance imaging can include receiving k-space data sets acquired by radiofrequency (RF) coils. Each of the k-space data sets can correspond to a different one of the RF coils. Each of the k-space data sets can be truncated and/or under sampled. The method can further include generating partial images of a field of view based on the k-space data sets and generating an initial image based on the partial images. The initial image can be full image of the field of view. The method can further include applying an iterative image reconstruction technique to generate an updated image based on the initial image.
1. A method for magnetic resonance imaging, the method comprising:
receiving k-space data sets acquired by radiofrequency (RF) coils of a RF coil assembly, wherein each of the k-space data sets correspond to a different one of the RF coils, and wherein each of the k-space data sets are truncated and under sampled;
receiving calibration k-space data sets acquired by the RF coils of the RF coil assembly, wherein each of the calibration k-space data sets correspond to a different one of the RF coils, and wherein each of the calibration k-space data sets comprises a central zone;
generate a phase map of the central zone based on the calibration k-space data sets;
generating partial images of a field of view based on the k-space data sets, wherein each of the partial images correspond to a different one of the k-space data sets;
generating an initial image based on the partial images, wherein the initial image is full image of the field of view; and
applying an iterative image reconstruction technique to generate an updated image based on the initial image, the iterative image reconstruction technique comprising:
designating the initial image as an input image for the iterative image reconstruction technique;
applying a phase correction to the input image to generate a first intermediate image;
applying a k-space conjugate synthesis to the input image to generate a second intermediate image;
calculating an output image based on the first intermediate image and the second intermediate image;
designating the output image as the input image for a next iteration; and
repeating the applying the phase correction to the input image, the applying the k-space conjugate synthesis to the input image, the calculating the output image, and the designating the output image as the input image for the next iteration, wherein the updated image is based on a final output image of the iterative image reconstruction technique.
2. A system, comprising:
an array of magnets configured to generate a low-field strength or ultra-low-field strength magnetic field toward an object of interest located within a field of view;
a radio frequency (RF) coil assembly comprising an array of RF coils, wherein the RF coils are positionable around an object of interest in the field of view, and wherein the RF coils are configured to acquire magnetic resonance signals; and
a control circuit comprising a processor and a memory, wherein the memory stores instructions executable by the processor to:
receive k-space data sets corresponding to magnetic resonance signals acquired by the RF coils, wherein each of the k-space data sets correspond to a different one of the RF coils, and wherein each of the k-space data sets are truncated and under sampled;
generate partial images of the field of view based on the k-space data sets, wherein each of the partial images correspond to a different one of the k-space data sets;
generate an initial image based on the partial images, wherein the initial image is full image of the field of view;
apply an iterative image reconstruction technique to generate an updated image based on the initial image;
wherein the instructions executable by the processor to apply the iterative image reconstruction technique to generate the updated image comprise instructions to:
designate the initial image as an input image for the iterative image reconstruction technique;
apply a phase correction to the input image to generate a first intermediate image;
apply a k-space conjugate synthesis to the input image to generate a second intermediate image;
calculate an output image based on the first intermediate image and the second intermediate image;
designate the output image as the input image for a next iteration; and
repeat the application of the phase correction to the input image, the application of the k-space conjugate synthesis to the input image, the calculation of the output image, and the designation of the output image as the input image for the next iteration until a difference between the output image and the corresponding input image satisfies a predetermined threshold, wherein the updated image is based on a final output image of the iterative image reconstruction technique.
3. The method of claim 1 , wherein the applying the phase correction to the input image, the applying the k-space conjugate synthesis to the input image, the calculating the output image, and the designating the output image as the input image for the next iteration is repeated until a difference between the output image and the corresponding input image satisfies a predetermined threshold.
4. The method of claim 1 , wherein applying the phase correction to the input image to generate the first intermediate image comprises:
determining a magnitude of the input image; and
calculating the first intermediate image based on the magnitude of the input image and the phase map of the central zone.
5. The method of claim 1 , wherein each of the k-space data sets are under sampled based on an under-sampling rate of at least 2 in a first transverse direction and a second transverse direction, and wherein each of the k-space data sets are truncated by at least 37.5% in the first transverse direction and the second transverse direction.
6. The method of claim 1 , wherein each of the k-space data sets are under sampled, truncated, and acquired in parallel such that a scan time required to acquire the k-space data sets is less than 10% of a scan time required to acquire a fully sampled, non-truncated k-space data set with a corresponding number of phase encodings.
7. The system of claim 2 , wherein the memory further stores instructions executable by the processor to:
receive calibration k-space data sets corresponding to magnetic resonance signals acquired by the RF coils, wherein each of the calibration k-space data sets correspond to a different one of the RF coils, and wherein each of the calibration k-space data sets comprises a central zone; and
generate a phase map of the central zone based on the calibration k-space data sets.
8. The system of claim 2 , wherein the k-space data sets comprise acquired k-space values, and wherein the instructions executable by the processor to apply the k-space conjugate synthesis to the input image to generate the second intermediate image comprise instructions to:
generate a first intermediate k-space by inverse Fourier transforming the input image, wherein the first intermediate k-space comprises intermediate k-space values;
generate a second intermediate k-space from the first intermediate k-space by replacing at least some of the intermediate k-space values with at least some of the acquired k-space values; and
generate the second intermediate image by Fourier transforming the second intermediate k-space.
9. The system of claim 2 , wherein the instructions executable by the processor to calculate the output image based on the first intermediate image and the second intermediate image comprise instructions to:
add a product of the first intermediate image and a first weighting factor to a product of the second intermediate image and a second weighting factor;
wherein a sum of the first weighting factor and the second weighting factor is equal to one.
10. The system of claim 2 , wherein each of the k-space data sets are under sampled based on an under-sampling rate of at least 2 in a first transverse direction and a second transverse direction, and wherein each of the k-space data sets are truncated by at least 37.5% in the first transverse direction and the second transverse direction.
11. The method of claim 4 , wherein the k-space data sets comprise acquired k-space values, and wherein applying the k-space conjugate synthesis to the input image to generate the second intermediate image comprises:
generating a first intermediate k-space by Fourier transforming the input image, wherein the first intermediate k-space comprises intermediate k-space values;
generating a second intermediate k-space from the first intermediate k-space by replacing at least some of the intermediate k-space values with at least some of the acquired k-space values; and
generating the second intermediate image by inverse Fourier transforming the second intermediate k-space.
12. The system of claim 7 , wherein the instructions executable by the processor to apply the phase correction to the input image to generate the first intermediate image comprise instructions to:
determine a magnitude of the input image; and
calculate the first intermediate image based on the magnitude of the input image and the phase map of the central zone.
13. The method of claim 11 , wherein calculating the output image based on the first intermediate image and the second intermediate image comprises:
adding a product of the first intermediate image and a first weighting factor to a product of the second intermediate image and a second weighting factor;
wherein a sum of the first weighting factor and the second weighting factor is equal to one.
14. The system of claim 12 , wherein the memory further stores instructions executable by the processor to generate a coil sensitivity map based on the calibration k-space data sets.
15. The method of claim 13 , further comprising generating a coil sensitivity map based on the calibration k-space data sets.
16. The system of claim 14 , wherein the instructions executable by the processor to generate the initial image based on the partial images comprise instructions to generate the initial image based on the partial images and the coil sensitivity map.
17. The method of claim 15 , wherein the k-space data sets are acquired in parallel, and wherein generating the initial image based on the partial images comprises generating the initial image based on the partial images and the coil sensitivity map.
18. The system of claim 16 , wherein the instructions executable by the processor to generate the initial image based on the partial images and the coil sensitivity map comprise instructions to generate the initial image according to a sensitivity encoding (SENSE) technique.
19. The method of claim 17 , wherein generating the initial image based on the partial images and the coil sensitivity map comprises generating the initial image according to a sensitivity encoding (SENSE) technique.