Sweeping real-time single point fiber

• US 6,859,203 B2
• Filed: 05/15/2002
• Issued: 02/22/2005
• Est. Priority Date: 05/15/2002
• Status: Expired due to Term

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Holthuizen, Ronaldus F. J., van Muiswinkel, Arianne M. C.
• Primary Examiner(s): Jankus, Almis R.
• Application Number:
  US10/145,917
• Publication Number:
  US 20030216634A1
• Time in Patent Office: 1,014 Days
• Field of Search: 345/419, 345/420, 345/424, 345/426, 345/427, 345/428, 324/307, 324/309, 324/312, 600/419
• US Class Current: 345/424
• CPC Class Codes: G01R 33/56341 Diffusion imaging
• CPC Class Codes: G06T 15/08 Volume rendering
• CPC Class Codes: G06T 2207/10092 Diffusion tensor magnetic resonance imaging [DTI]
• CPC Class Codes: G06T 2207/20221 Image fusion; Image merging
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 5/50 using two or more images, e.g. averaging or subtraction

Abstract

An imaging method for imaging a subject (18) including fibrous/anisotropic structures (102) includes acquiring three-dimensional image representations without and with a plurality of different diffusion weighting and directions. When a user (56) hovers a selection device over a voxel of the image, a fiber representation (54) is extracted in substantially real time. The representation is generated by following a direction of a major eigenvector e₁ from voxel to voxel. A human-viewable display of the fiber representation is produced (210).

Claims

1. A method of diffusion imaging comprising:

a) generating an electronic static image representation and a plurality of electronic diffusion weighted image representations;

b) converting one of the images to a human viewable display;

c) selecting a single starting voxel on the human readable display;

d) from the electronic image representations, determining a direction of a maximum diffusion eigenvalue in the starting voxel;

e) identifying a single neighboring voxel in the determined direction of maximum diffusion;

f) from the electronic image representations, determining a direction of maximum diffusion in the neighboring voxel;

g) repeating steps e) and f) to generate a single fiber path through neighboring voxels that represents a single fiber maximum diffusion flow path;

h) converting the single fiber path to a human viewable display.

2. The method as set forth in claim 1, wherein the single fiber path is generated and converted into a human readable display substantially in real time.

3. The method as set forth in claim 1, wherein the converting step b) includes:

colorizing the displayed images.

4. The method as set forth in claim 1, wherein the selecting step c) includes:

hovering a selection cursor over the human readable display to display a series of single fiber paths, one through each selected voxel.

5. The method as set forth in claim 1, wherein the determining the direction of maximum diffusion step includes:

finding major, medium, and minor eigenvectors of the voxel;

identifying the major eigenvector.

6. The method as set forth in claim 1, further including:

in step e) identifying neighboring voxels in both positive and negative directions of a major eigenvector.

7. The method as set forth in claim 1, further including:

repeating step g) until a termination criterion is met.

9. The method as set forth in claim 1, wherein the converting step h) includes:

superimposing the fiber path on the display of step b).

10. The method as set forth in claim 1, wherein the step of determining the direction of maximum diffusion includes:

determining an apparent diffusion coefficient tensor from the static and diffusion weighted image representations;

extracting eigenvectors and eigenvalues from the apparent diffusion coefficient tensor.

11. The method as set forth in claim 10, further including:

from the eigenvectors and eigenvalues, computing an anisotropy value for each voxel.

12. The method as set forth claim 11, wherein the step h) includes:

rendering an image of the fractional anisotropy of the voxels of the fiber path;

colorizing the rendered image.

8. A method of diffusion imaging comprising:

a) generating an electronic static image representation and a plurality of electronic diffusion weighted image representations;

b) converting one of the images to a human viewable display;

c) selecting a starting voxel on the human readable display;

d) from the electronic image representations, determining a direction of a maximum diffusion eigenvalue in the starting voxel;

e) identifying a neighboring voxel in the determined direction of maximum diffusion;

f) from the electronic image representations, determining a direction of maximum diffusion in the neighboring voxel;

g) repeating steps e) and f) to generate a track of neighboring voxels;

h) repeating step g) until at least one of the following termination criteria is met;

encountering a voxel with a major Eigenvector not greater than a minor or medium Eigenvector of the same voxel by a predetermined threshold;

encountering a voxel with a major Eigenvector direction different from a previous Eigenvector direction by greater than a predetermined angle;

exceeding a timeout threshold;

encountering a voxel without diffusion weighting low a signal intensity threshold; and

i) converting the track to a human viewable display.

13. A magnetic resonance imaging apparatus comprising:

a gantry which acquires a plurality of image representations without and with various diffusion weighting and directions;

a monitor which converts one of the image representations to a human readable display;

a manual input device through which an operator selects one of a voxel and a group of voxels; and

an image processor programmed to;

calculate a path of a fiber center line that intersects each selected one of the selected voxels;

overlay an image of each fiber center line onto the human readable display on the monitor.

14. A magnetic resonance apparatus comprising:

a means for acquiring a plurality of image representation without and with each of a plurality of diffusion weighting and directions;

a display means for converting the image representations to a human readable image;

a means for selecting at least one voxel as a starting point;

means for determining a direction of a maximum diffusion eigenvalue in the starting voxel;

means for identifying a voxel neighboring the starting point in the determined direction of maximum diffusion;

a means for determining a direction of maximum diffusion in the neighboring voxel;

a means for calculating a path of fluid diffusing along a fiber that intersects the starting point by causing the neighboring voxel identifying means and the neighboring voxel maximum diffusion direction calculating means to generate a track of neighboring voxels until a least one of the following terminate on criteria is met;

encountering a voxel with a major Eigenvector not greater than a minor or medium Eigenvector of the same voxel by pre-determined threshold;

encountering a voxel with a major Eigenvector direction different from a previous Eigenvector direction by greater than a pre-determined angle;

exceeding a timeout threshold;

encountering a voxel without diffusion weighting below a signal intensity threshold; and

a means for overlaying the calculated fiber bundle path on the human readable image.

15. The magnetic resonance apparatus as set forth in claim 14, wherein the means for selecting includes:

a manually controlled cursor displayed on the display means.

20. The magnetic resonance apparatus as set forth in claim 14, further including:

a means for calculating an apparent diffusion coefficient;

a means for calculating eigenvectors.

21. The magnetic resonance apparatus as set forth in claim 14, further including:

a means for calculating and displaying various anisotropy maps.

16. A magnetic resonance apparatus comprising:

a means for acquiring a plurality of image representation without and with each of a plurality of diffusion weighting and directions;

a display means for converting the image representations to a human readable image;

a means for selecting one of a voxel d a group of voxel as a starting point that includes a manually controlled cursor displayed on the display means;

a calculating means that interpolates the voxels giving them a higher resolution than acquired voxel and calculates a path of fluid diffusing along a fiber bundle that intersects the starting point; and

the calculating means including;

a diffusion direction processing means that;

a) identifies a major Eigenvector of the starting point;

b) identifies an adjacent voxel by following a direction of the major Eigenvector;

c) finds a diffusion direction of the adjacent voxel by identifying a major Eigenvector of the adjacent voxel;

d) marks the adjacent voxel;

e) repeats steps b), c), and d) until a termination criterion is met;

a means for overlaying the calculated fiber bundle path on the human readable image.

17. The magnetic resonance apparatus as set forth in claim 16, wherein the diffusion direction processing means iteratively performs steps b), c), and d) in a negative direction of the major eigenvector.

18. The magnetic resonance apparatus as set forth in claim 16, further including:

a termination criterion recognizing mean that recognizes at least one of;

encountering a voxel with a major eigenvector not greater than a minor or medium eigenvector of the same voxel by a predetermined threshold;

encountering a voxel with a major eigenvector direction different from a previous eigenvector direct on by greater than predetermined angle;

exceeding a time threshold.

19. The magnetic resonance apparatus as set forth in claim 18, wherein the major, medium, and minor eigenvectors are mutually orthogonal vectors.

22. A method of displaying a diffusion pathway comprising:

gathering a diffusion weighted, multi-direction, three-dimensional magnetic resonance image representation of a region of a subject containing fibrous structures;

selecting a starting point within the image representation;

calculating eigenvectors for voxels in the image representation along a center line of a flow path which intersects a starting point;

forming from the magnetic resonance image representation a combined image representation and contrasted voxels along the center line of the flow path that intersects the selected starting point;

displaying the combined image representation in a human readable format.

23. A method of magnetic resonance comprising:

acquiring a plurality of image representations without and with each of a plurality of diffusion weighting and directions;

converting the image representations into a human readable image;

selecting one of a voxel and a group of voxels as a starting point;

interpolating the voxels giving them a higher resolution than the voxels of the acquired images;

calculating a path of fluid diffusing along a fiber bundle that intersects the starting point; and

,

overlaying the calculated fiber bundle path on the human readable image.