Magnetic resonance imaging apparatus and magnetic resonance imaging method

1. A magnetic resonance imaging (MRI) apparatus comprising:

MRI system components including static and gradient magnetic field generators, radio frequency (RF) transmitter and receiver circuits, at least one RF coil and at least one processor connected to control said components as to

acquire first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generate first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generate second magnetic resonance k-space image data, based on the first amplitude image data, by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data;

in response to transform data processing of the amplitude image data being performed, perform filtering of the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being used, employing a filter function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein the window function H homo (k) has a gain in a high frequency region and again in a low frequency region, the gain in the high frequency region being larger than the gain in the low frequency region, the high frequency region consisting of the non-sampled region and an asymmetrically sampled portion out of the asymmetrically sampled region, the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region.

2. The magnetic resonance imaging apparatus of claim 1 , wherein the gain in the high frequency region of the filter function H homo (k) is not more than twice the gain in the low frequency region.

3. The magnetic resonance imaging apparatus of claim 1 , wherein the window function H homo (k) is a smoothly varying function.

4. A magnetic resonance imaging apparatus comprising:

MRI system components including static and gradient magnetic field generators, radio frequency (RF) transmitter and receiver circuits, at least one RF coil and at least one processor connected to control said components as to

acquire first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generate first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generate second magnetic resonance k-space image data by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data; and

in response to transform data processing of the amplitude image data being performed, perform filtering of the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being performed, employing a function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein the window function H homo (k) has a gain in a high frequency region and again in a low frequency region, the gain in the high frequency region being larger than the gain in the low frequency region, the high frequency region being an asymmetrically sampled portion out of the asymmetrically sampled region, the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region.

5. The magnetic resonance imaging apparatus of claim 4 , wherein the gain of filter function H homo (k) in the high frequency region is not less than twice the gain in the low frequency region and not more than four times the gain in the low frequency region, and a gain in the non-sampling region is zero.

6. A magnetic resonance imaging apparatus comprising:

MRI system components including static and gradient magnetic field generators, radio frequency (RF) transmitter and receiver circuits, at least one RF coil and at least one processor connected to control said components as to

acquire first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generate first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generate second magnetic resonance k-space image data by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data; and

in response to transform data processing of the amplitude image data being performed, perform filtering of the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being performed, employing a filter function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein said at least one computer is configured to perform a series of processing one time or plural times, the series including first processing, second processing, third processing, fourth processing, fifth processing, and sixth processing,

the first processing being performed for the magnetic resonance image data and extracting a real part of real space data,

the second processing shifting a phase of the real part in a direction opposite to a phase correction direction,

the third processing converting real space data after the second processing to k-space data,

the fourth processing replacing a part of the k-space data derived by the third processing, the part corresponding to the sampled region, being replaced with magnetic resonance data in the sampled region,

the fifth processing converting k-space data derived by the fourth processing to real space data, and

the sixth processing shifting a phase of the real space data derived by the fifth processing in the phase correction direction to generate updated magnetic resonance image data.

7. The magnetic resonance imaging apparatus of claim 6 , wherein

the second processing is performed using a phase distribution in a low frequency region and the sixth processing using a phase distribution corresponding to the sampling region, the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region,

the phase distribution in the low frequency region being derived based on real space data derived by converting magnetic resonance data in the low frequency region out of the magnetic resonance data in the asymmetrically sampled region, and

the phase distribution corresponding to the sampled region being derived based on real space data derived by converting the magnetic resonance data in the asymmetrically sampled region.

8. A magnetic resonance imaging (MRI) method comprising:

acquiring first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generating first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generating second magnetic resonance k-space image data by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data; and

in response to transform data processing of the amplitude image data being performed, filtering the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being performed, employing a filter function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein the window function H homo (k) has a gain in a high frequency region and again in a low frequency region, the gain in the high frequency region being larger than the gain in the low frequency region, the high frequency region consisting of the non-sampled region and an asymmetrically sampled portion out of the asymmetrically sampled region, the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region.

9. The magnetic resonance imaging method of claim 8 , wherein the gain of said filter function H homo (k) in the high frequency region is not more than twice the gain in the low frequency region.

10. The magnetic resonance imaging method of claim 8 , wherein the window function H homo (k) is a smoothly varying function.

11. A magnetic resonance imaging method comprising:

acquiring first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generating first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generating second magnetic resonance k-space image data by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data; and

in response to transform data processing of the amplitude image data being performed, filtering the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being performed, employing a filter function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein the window function H homo (k) has a gain in a high frequency region and again in a low frequency region, the gain in the high frequency region being larger than the gain in the low frequency region, the high frequency region being an asymmetrically sampled portion out of the asymmetrically sampled region, the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region.

12. The magnetic resonance imaging method of claim 11 , wherein the gain of said filter function H homo (k) in the high frequency region is not less than twice the gain in the low frequency region and not more than four times the gain in the low frequency region, and a gain in the non-sampling region is zero.

13. A magnetic resonance imaging apparatus comprising:

acquiring first magnetic resonance k-space data from an object, the magnetic resonance data corresponding to a sampling region asymmetric in a wave number direction in k-space;

generating first amplitude image data, in real space, based on first k-space data after zero padding to a non-sampled region of the magnetic resonance data and without phase correction of the first k-space data, said zero padding being performed by applying an H low (k) window function which defines a symmetrical portion of the acquired magnetic resonance data in a low frequency region from the acquired asymmetric MR data, where zero padding is performed outside the H low (k) window function region;

generating second magnetic resonance k-space image data by (a) transform data processing of the amplitude image data or (b) convolution processing of the amplitude image data, wherein phase components of the generated second k-space image data simulate phase-corrected MoFIR (modified finite impulse response) k-space data; and

in response to transform data processing of the amplitude image data being performed, filtering the second k-space data by a window function H homo (k) and then converting the filtered second k-space data into second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR, or

in response to convolution processing of the amplitude image data being performed, employing a filter function in real space derived by converting said window function H homo (k) to produce second real space image data with reduced deterioration in high frequency components simulating real space image data generated by MoFIR,

wherein said at least one computer is configured to perform a series of processing one time or plural times, the series including first processing, second processing, third processing, fourth processing, fifth processing, and sixth processing,

the first processing being performed for the magnetic resonance image data and extracting a real part of real space data,

the second processing shifting a phase of the real part in a direction opposite to a phase correction direction,

the third processing converting real space data after the second processing to k-space data,

the fourth processing replacing a part of the k-space data derived by the third processing, the part corresponding to the sampled region, being replaced with magnetic resonance data in the sampled region,

the fifth processing converting k-space data derived by the fourth processing to real space data, and

the sixth processing shifting a phase of the real space data derived by the fifth processing in the phase correction direction to generate updated magnetic resonance image data.

14. The magnetic resonance imaging method of claim 13 , wherein

the second processing is performed using a phase distribution in a low frequency region and the sixth processing using a phase distribution corresponding to the sampling region,

the low frequency region being a symmetrically sampled portion out of the asymmetrically sampled region,

the phase distribution in the low frequency region being derived based on real space data derived by converting magnetic resonance data in the low frequency region out of the magnetic resonance data in the asymmetrically sampled region, and

the phase distribution corresponding to the sampled region being derived based on real space data derived by converting the magnetic resonance data in the asymmetrically sampled region.