Method for reconstructing images of a multi-channel MRI system

1. A method for reconstructing a an image with a magnetic resonance image (MRI) system that includes a radio frequency (RF) coil array having multiple coil channels, the method comprising:

a. measuring a first set of complex-valued k-space data at each channel of a multi-channel coil array;

b. reinforcing the linear relationship between each data point of the first set of k-space data in one channel of the coil array and its neighboring k-space data points in other channels of the coil array to generate a second set of k-space data; and

c. reconstructing a magnetic resonance image using the second set of k-space data;

wherein said reconstructing the magnetic resonance image in step c. is calculated based on a data consistency term associated with the first set of complex-valued k-space data and a measurement consistency term associated with the second set of k-space data.

2. The method according to claim 1 , wherein step b. is achieved by a procedure comprising:

using the first set of k-space data at each channel of the multi-channel coil array to estimate a set of linear combination coefficients, which describes the linear relationship between each data point of the first set of k-space data in one channel of the coil array and its neighboring k-space data points in other channels of the coil array;

generating the second set of k-space data such that the second set of k-space data is simultaneously similar to the first set of k-space data and to a set of synthesized k-space data wherein each data point of the set of synthesized k-space data is the linear combination of its neighboring k-space data points using the set of linear combination coefficients.

3. The method according to claim 2 , wherein estimating the set of linear combination coefficients uses the third set of k-space data collected in a separate MRI measurement using the same RF coil array.

4. The method according to claim 2 , wherein estimating the set of linear combination coefficients is based on least squares fitting.

5. The method according to claim 2 , wherein the similarity between the first and the second sets of the k-space data and the similarity between the second and the set of synthesized k-space data is based on the sum of squares of the difference.

6. The method according to claim 2 , wherein the similarity between the first set of k-space data and the second set of k-space data can be different from the similarity between the set of synthesized k-space data and the second set of k-space data.

7. The method according to claim 2 , wherein the second set of k-space data can be generated by an iterative calculation procedure.

8. The method according to claim 2 , wherein step b. is achieved by a procedure comprising:

using the first set of k-space data at each channel of the multi-channel coil array to estimate a set of linear combination coefficients, which describes the linear relationship between each data point of the first set of k-space data in one channel of the coil array and its neighboring k-space data points in other channels of the coil array;

generating the second set of k-space data such that the second set of k-space data is simultaneously similar to the first set of k-space data, similar to a set of synthesized k-space data wherein each data point of the set of synthesized k-space data is the linear combination of its neighboring k-space data points using the set of linear combination coefficients, and the second set of k-space data is sparse after applying the Fourier transform to the second set of k-space data to generate an image.

9. The method according to claim 8 , wherein estimating the set of linear combination coefficients uses the third set of k-space data collected in a separate MRI measurement using the same RF coil array.

10. The method according to claim 8 , wherein estimating the set of linear combination coefficients is based on least squares fitting.

11. The method according to claim 8 , wherein the similarity between the first and the second sets of the k-space data and the similarity between the second and the set of synthesized k-space data is based on the sum of squares of the difference.

12. The method according to claim 8 , wherein the similarity between the first set of k-space data and the second set of k-space data can be different from the similarity between the set of synthesized k-space data and the second set of k-space data.

13. The method according to claim 8 , wherein the second set of k-space data can be generated by an iterative calculation procedure.

14. The method according to claim 8 , wherein the similarity between the first set of k-space data and the second set of k-space data, the similarity between the set of synthesized k-space and the second set of k-space data, and the sparsity of the second set of k-space data after applying the Fourier transform to the second set of k-space data to generate the image can be different.

15. The method according to claim 8 , wherein the sparsity of the second set of k-space data after applying the Fourier transform to the second set of k-space data to generate the image is measured by the L-p norm of the transformed image, where p is between 0 and 1.

16. The method according to claim 15 , wherein the sparsity of the second set of k-space data after applying the Fourier transform to the second set of k-space data to generate an image can measured by wavelet transformation, Total variation transformation, or Laplacian transformation.

17. The method according to claim 1 , wherein the data consistency term is calculated including a convolution operation using a plurality of weightings over the first set of complex-valued k-space data, while the measurement consistency term is calculated based at least on a k-space sampling matrix and an acquired data across the channels of the multi-channel coil array.

18. The method according to claim 17 , wherein the data consistency term is a calculation associated with |Ad−d| where A is associated with a matrix consisting of convolution kernel and d is associated with the vertical concatenation of the complex-valued k-space data, while the measurement consistency term is a calculation associated with |Ed−d 0 | where E is associated with the k-space sampling matrix; d 0 is associated with the acquired data across the channels of the multi-channel coil array.