Artifact suppression in dynamic magnetic resonance imaging

• US 20020156364A1
• Filed: 04/19/2002
• Published: 10/24/2002
• Est. Priority Date: 04/20/2001
• Status: Active Grant

• Current Assignee: Brigham and Women's Hospital Incorporated
• Original Assignee: Brigham and Women's Hospital Incorporated
• Inventors: Madore, Bruno
• Granted Patent: US 7,283,859 B2
• Time in Patent Office: Days
• US Class Current: 600/410
• CPC Class Codes: G01R 33/5611 Parallel magnetic resonance imaging, e.g. sensitivity encoding [SENSE], simultaneous acquisition of spatial harmonics [SMASH], unaliasing by Fourier encoding of the overlaps using the temporal dimension [UNFOLD], k-t-broad-use linear acquisition speed-up technique [k-t-BLAST], k-t-SENSE structural details of arrays of sub-coils G01R33/3415
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/10088 Magnetic resonance imaging [MRI]
• CPC Class Codes: G06T 2207/20056 Discrete and fast Fourier transform, [DFT, FFT]
• CPC Class Codes: G06T 5/10 using non-spatial domain filtering
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction
• CPC Class Codes: G06T 5/70 Denoising; Smoothing

Abstract

A method for removing one or more artifacts from a time series of magnetic resonance (MR) images is provided. In one embodiment of the invention, an MR time series of data sets is acquired while changing the k-space that is sampled at each time frame. By using the acquired data, information is produced for a plurality of MR images at different time points, wherein the images contain desired components and one or more artifacts. Changing the k-space locations that are sampled from time frame to time frame makes the artifacts behave in a signature way through time, so that they can be identified as being artifacts and be removed through temporal analysis.

Claims

1. A method for removing one or more leftover artifacts from a time series of dynamic magnetic resonance (MR) images, the method comprising:

acquiring an MR data set by changing k-space locations that are sampled at each time point;

separating desired components from one or more leftover artifacts by filtering the one or more leftover artifacts from the desired components in a temporal frequency domain;

producing information for a first MR image at a first time point using a first MR reconstruction technique; and

producing information for a second MR image at a second time point using the first MR reconstruction technique.

2. The method of claim 1, wherein separating the desired components from the one or more leftover artifacts comprises:

(a) Fourier-transforming an image or k-space pixel over time to obtain a result of a Fourier transform;

(b) filtering the result of the Fourier transform to separate the desired components from the one or more leftover artifacts; and

(c) repeating steps (a) and (b) above for a plurality of image or k-space pixels.

3. A method of claim 2, wherein filtering the result of the Fourier transform comprises removing a frequency band that is located around a Nyquist frequency.

4. A method of claim 1, wherein the MR data set has one or more artifacts centered at a DC frequency in the temporal frequency domain, the method further comprising:

producing information for a third MR image at the first time point using a second MR reconstruction technique on the MR data set; and

producing information for a fourth MR image at the second time point using the second MR reconstruction technique on the MR data set;

wherein separating desired components from one or more leftover artifacts comprises;

replacing a frequency band around the DC frequency in information treated with the first MR reconstruction technique with a corresponding frequency band from information treated with the second MR reconstruction technique.

5. The method of claim 1, wherein acquiring an MR data set by changing the k-space that is sampled at each time point comprises

sampling a first set of k-space locations every odd frame and sampling a second set of k-space locations every even frame with a modified partial Fourier imaging technique.

6. The method of claim 1, wherein acquiring an MR data set by changing the k-space that is sampled at each time point comprises:

sampling a first set of k-space locations every odd frame; and

sampling a second set of k-space locations every even frame.

7. The method of claim 1, wherein the MR reconstruction technique is SENSE.

8. The method of claim 1, wherein the MR reconstruction technique is SMASH.

9. The method of claim 1, further comprising converting noise amplification into an artifact such that the artifact can be reduced.

10. A method for removing one or more leftover artifacts from a time series of dynamic magnetic resonance (MR) images, the method comprising:

obtaining a first set of information about an object at a first time point and at a first set of k-space locations;

obtaining a second set of information about the object at a second time point and at a second set of k-space locations, wherein at least one of the k-space locations in the second set of information is not contained in the first set of information;

producing data for first and second images by using a first reconstruction technique on the first and the second sets of information, respectively;

producing data for third and fourth images using a second reconstruction technique on the first and the second sets of information, respectively, wherein the second reconstruction technique is different than the first reconstruction technique;

removing one or more frequency bands from data treated with the first reconstruction technique, in a temporal frequency domain; and

replacing one or more of the removed frequency bands with corresponding frequency bands obtained from data treated with the second reconstruction technique.

11. The method of claim 10, wherein removing one or more frequency bands comprises:

Fourier-transforming a time series of image or k-space pixels from the data treated with the first MR reconstruction technique; and

filtering the result of the Fourier transform to remove frequency bands.

12. The method of claim 10, wherein removing a frequency band comprises removing in a temporal frequency domain a frequency band at a DC frequency in the data treated with the first MR reconstruction technique.

13. The method of claim 12, further comprising removing in a temporal frequency domain a frequency band near a Nyquist frequency in the data treated with the first MR reconstruction technique.

14. The method of claim 10, further comprising removing in a temporal frequency domain a frequency band near a Nyquist frequency in the data treated with the first MR reconstruction technique.

15. The method of claim 10, wherein the first reconstruction technique is a parallel imaging algorithm with a predetermined acceleration factor, and the second reconstruction technique is a parallel imaging algorithm with an acceleration factor twice smaller than the acceleration factor used in the first reconstruction technique.

16. The method of claim 10, further comprising converting noise amplification into an artifact such that the artifact can be reduced.

18. The method of claim 17, further comprising removing from the first set of image data a second frequency band being centered near a Nyquist frequency in the temporal frequency domain.

17. A method for removing one or more leftover artifacts from a plurality of dynamic magnetic resonance (MR) images, the method comprising:

obtaining a first set of information about the object at a first time point and at a first set of k-space locations;

obtaining a second set of information about the object at a second time point and at a second set of k-space locations, wherein at least one of the k-space locations in the second set of information is not contained in the first set of information;

producing a first set of image data at the first and second time points using a first MR reconstruction method on the first and the second sets of information, the first set of image data having at least one artifact centered at DC;

producing a second set of image data at the first and second time points using a second MR reconstruction method on the first and the second sets of information, the second set of image data being substantially artifact free at DC;

removing from the first set of image data a frequency band being centered at DC in the temporal frequency domain; and

replacing the frequency band of the first set of image data with a corresponding frequency band centered at DC of the second set of image data.

19. An article of manufacture, comprising:

a computer usable medium having computer readable program code means embodied therein for removing one or more leftover artifacts from a time series of dynamic magnetic resonance (MR) images, the computer readable program code means in said article of manufacture comprising;

computer readable program code means for causing the computer system to acquire an MR data set by changing k-space locations that are sampled at each time point;

computer readable program code means for causing the computer system to separate desired components from one or more leftover artifacts by filtering the one or more leftover artifacts from the desired components in a temporal frequency domain;

computer readable program code means for causing the computer system to produce information for a first MR image at a first time point using a first MR reconstruction technique; and

computer readable program code means for causing the computer system to produce information for a second MR image at a second time point using the first MR reconstruction technique.

20. An article of manufacture, comprising:

a computer usable medium having computer readable program code means embodied therein for removing one or more leftover artifacts from a time series of dynamic magnetic resonance (MR) images, the computer readable program code means in said article of manufacture comprising;

computer readable program code means for causing the computer system to obtain a first set of information about an object at a first time point and at a first set of k-space locations;

computer readable program code means for causing the computer system to obtain a second set of information about the object at a second time point and at a second set of k-space locations, wherein at least one of the k-space locations in the second set of information is not contained in the first set of information;

computer readable program code means for causing the computer system to produce data for first and second images by using a first reconstruction technique on the first and the second sets of information, respectively;

computer readable program code means for causing the computer system to produce data for third and fourth images using a second reconstruction technique on the first and the second sets of information, respectively, wherein the second reconstruction technique is different than the first reconstruction technique;

computer readable program code means for causing the computer system to remove one or more frequency bands from the data treated with first reconstruction technique, in a temporal frequency domain; and

computer readable program code means for causing the computer system to replace one or more of the removed frequency bands with corresponding frequency bands obtained from the data treated with the second reconstruction technique.