fMRI signal processing

• US 6,370,416 B1
• Filed: 11/25/1998
• Issued: 04/09/2002
• Est. Priority Date: 11/25/1998
• Status: Expired due to Term

• Current Assignee: GE Medical Systems Global Technology Company LLC (GE Healthcare Technologies, Inc.)
• Original Assignee: GE Medical Systems Global Technology Company LLC (GE Healthcare Technologies, Inc.)
• Inventors: Rosenfeld, Daniel
• Primary Examiner(s): Lateef, Marvin M.
• Assistant Examiner(s): SHAW, SHAWNA JEANNINE
• Application Number:
  US09/199,326
• Time in Patent Office: 1,231 Days
• Field of Search: 600/410, 600/407, 600/544, 324/309, 382/128, 382/254, 382/275, 128/920, 128/922
• US Class Current: 600/410
• CPC Class Codes: G01R 33/56 Image enhancement or correction, e.g. subtraction or averaging techniques , e.g. improvement of signal-to-noise ratio and resolution
• CPC Class Codes: G01R 33/5608 Data processing and visualization specially adapted for MR, e.g. for feature analysis and pattern recognition on the basis of measured MR data, segmentation of measured MR data, edge contour detection on the basis of measured MR data, for enhancing measured MR data in terms of signal-to-noise ratio by means of noise filtering or apodization, for enhancing measured MR data in terms of resolution by means for deblurring, windowing, zero filling, or generation of gray-scaled images, colour-coded images or images displaying vectors instead of pixels image data processing or generation, in general G06T
• CPC Class Codes: Y10S 128/922 including image analysis

Abstract

A method for MRI signal analysis of an area of a biological tissue comprising:

a) providing a biological tissue, wherein physiological activity is taking place in an area thereof;

b) acquiring sequential magnetic resonance images, at least during a portion of time in which said physiological activity is taking place, of said area and of at least a portion of the tissue in a vicinity of the area;

c) constructing, responsive to at least one pixel-related parameter value of said images, a pixel parameter space; and

d) separating the pixel parameter space into at least two subspaces.

Claims

1. A method for functional MRI (fMRI) signal analysis of an area of the brain comprising:

a) invoking neuronal activity in an area of the brain of a subject;

b) acquiring sequential magnetic resonance images, at least during a portion of time in which said are of the brain of a subject is activated, of said area and of at least a portion of the brain in a vicinity of the area;

c) constructing, responsive to at least one pixel-related parameter value of said images, a pixel parameter space; and

d) applying non-linear filtering to said pixel parameter space, wherein said nonlinear filtering separates said pixel parameter space into at least two subspaces, comprising a signal subspace and a noise subspace,

the method further comprising deriving a plurality of detection functions from image portions of a region of the brain in the vicinity of the area in said sequential images; and

synchronously detecting temporal variations in said area, over a sequence of images, using synchronous detection.

2. A method according to claim 1 wherein said at least one pixel-related parameter comprises a pixel magnitude value.

3. A method according to claim 2 and comprising arranging said magnitude values in a matrix.

4. A method according to claim 3 and comprising decomposing said matrix into a product of a plurality of matrices, the columns of one matrix of the plurality of matrices comprising a plurality of basis vectors which span said at least two subspaces.

5. A method according to claim 4 wherein said at least two subspaces comprise a signal subspace and a noise subspace.

6. A method according to claim 5 wherein one basis vector spans said signal subspace.

7. A method according to claim 6 wherein said basis vector is used as a detection function.

8. A method according to claim 7, wherein said detection comprises synchronous detection.

9. A method according to claim 5 wherein two or more basis vectors span said signal subspace.

10. A method according to claim 9 wherein said basis vectors are used as synchronous detection functions.

11. A method according to claim 1 wherein said non-linear filtering uses singular value decomposition.

12. A method according to claim 1 comprising detecting temporal variations in said area over a sequence of images using a synchronous detection function.

13. A method according to claim 12 and further comprising:

a) repeating said synchronous detection on all the pixels related to an area of interest in the brain;

b) obtaining, for each said synchronous detection, at least one detection indicator which indicates a quality of said synchronous detection; and

c) identifying those of said at least one detection indicator which meet at least one certain criterion.

14. A method according to claim 1 further comprising:

a) obtaining a plurality of detection indicator values from said synchronous detection, for a plurality of pixels related to an area of interest in the brain;

b) constructing a vector from detection indicator values associated with a single pixel, for each pixel of said plurality of pixels;

c) calculating the magnitude of each of said vectors; and

d) identifying those of said magnitudes which meet at least one certain criterion.

15. A method according to claim 14 further comprising deriving a vector phase from each of said vectors.

16. A method according to claim 15, comprising detecting a delay between the neuronal activity and a cause which invoked it using said phase of each vector.

17. A method according to claim 14 comprising determining at least a relative intensity of said neuronal activity from said magnitude of each vector.

18. A method according to claim 14 comprising determining an absolute intensity of said neuronal activity from said magnitude of each vector.

19. A method according to claim 1 wherein the neuronal activity in said area of the brain of a subject is periodic.

20. A method according to claim 1 wherein at least one of said MR images is acquired after a cession of said invoking of neuronal activity.

21. A method according to claim 1 comprising removing trends unrelated to said neuronal activity before said applying non linear filtering.

22. A method according to claim 1 comprising:

a) invoking at least a second activity simultaneous with the first activity;

b) independently extracting signals related to said at least two invoked activities.

23. A method for functional MRI (fMRI) signal analysis of an area of the brain comprising:

a) invoking neuronal activity in an area of the brain of a subject;

b) acquiring sequential magnetic resonance images, at least during a portion of time in which said area of the brain of a subject is activated, of said area and of at least a portion of the brain in a vicinity of the area;

c) registering said sequential images;

d) constructing a pixel parameter space, responsive to pixel magnitude values in said images;

e) removing, from said space, at least one trend unrelated to said neuronal activity;

f) arranging said trend-removed magnitude values in a matrix;

g) separating said pixel parameter space into at least two subspace, using singular value decomposition, such that said matrix is decomposed into a product of a plurality of matrices, the columns of one matrix of the plurality of matrices comprising a plurality of basis victors which span said at least two subspaces;

h) identifying said at least two subspaces as a signal subspace and as a noise subspace;

i) selecting at least one of said basis vectors which spans said signal subspace;

j) detecting variations in said area of the brain over a sequence of images using said at least one vector as a synchronous detection function, for all the pixels related to an area of interest in the brain;

k) obtaining, for each said synchronous detection, at least one detection indicator which indicates a quality of said synchronous detection.

24. A method according to claim 23, wherein said at least one basis vector which spans said signal subspace comprises at least two independent basis vectors.

25. A method according to claim 24, wherein a plurality of detection indicators are obtained for each pixel, comprising:

constructing a vector from said detection indicators; and

calculating the magnitude of each said vector.