Method and apparatus for evaluating regional changes in three-dimensional tomographic images

• US 8,280,482 B2
• Filed: 04/19/2005
• Issued: 10/02/2012
• Est. Priority Date: 04/19/2004
• Status: Active Grant

• Current Assignee: New York University
• Original Assignee: New York University
• Inventors: Rusinek, Henry, De Leon, Mony J.
• Primary Examiner(s): Casler, Brian
• Assistant Examiner(s): Shahrestani, Nasir
• Application Number:
  US11/109,340
• Publication Number:
  US 20050244036A1
• Time in Patent Office: 2,723 Days
• Field of Search: 600/407, 600/476, 600/443, 600/410, 600/447, 382/128, 382/173
• US Class Current: 600/407
• CPC Class Codes: A61B 5/05 Detecting, measuring or recording for diagnosis by means of electric currents or magnetic fields; Measuring using microwaves or radio waves impedance plethysmography A61B5/0295; measuring movement of the entire body or parts thereof A61B5/11; detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof A61B5/24
• CPC Class Codes: G06T 2207/30016 Brain
• CPC Class Codes: G06T 7/0012 Biomedical image inspection

Abstract

Methods for measuring atrophy in a brain region occupied by the hippocampus and entorhinal cortex. In one example, MRI scans and a computational formula are used to measure the medial-temporal lobe region of the brain over a time interval. This region contains the hippocampus and the entorhinal cortex, structures allied with learning and memory. Each year this region of the brain shrank in people who developed memory problems up to six years after their first MRI scan. The method is also applicable for measuring the progression rate of atrophy in the region in an instance where the onset of Alzheimer'"'"'s disease has already been established.

Claims

1. A method for determining a change in at least one portion of a biological object, comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the object;

determining a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image;

determining a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein;

using a computer arrangement, normalizing the first region of interest by determining a normalization factor, and adjusting a signal associated with one of the first image or the second image by the normalization factor; and

determining the change in the at least one portion by comparing the normalized first region of interest to the second region of interest.

2. The method of claim 1, further comprising the operation of segmenting the first region of interest.

3. The method of claim 1, wherein the operation of normalizing the first region of interest comprises:

determining a difference in a value between the first region of interest and the second region of interest; and

adjusting the value of the first region of interest accordingly.

4. The method of claim 3, wherein the value is selected from at least one of a signal intensity or contrast.

5. The method of claim 3, wherein the operation of normalizing the first region of interest is carried out without reference to a portion of the at least one of the first three-dimensional image or the second three-dimensional image outside the region of interest.

6. The method of claim 1, wherein:

the first region of interest is determined within the first image;

the second region of interest is determined within the second image; and

the first and second regions of interest correspond when the first and second images are coregistered.

7. The method of claim 1, wherein the operation of adjusting the signal associated with one of the first image or the second image is applied only across the first region of interest.

8. The method of claim 6, further comprising the operations of:

computing a first partial volume for the first region of interest;

computing a second partial volume for the second region of interest; and

determining the change between the first and second partial volumes.

9. The method of claim 8, further comprising the operations of:

reducing a first noise associated with the first partial volume prior to determining the change between the first and second partial volumes; and

reducing a second noise associated with the second partial volume prior to determining the change between the first and second partial volumes.

12. The method of claim 1, further comprising the operations of:

determining a first voxel set contained within the first region of interest;

determining a second voxel set contained within the second region of interest;

operating on the first voxel set and second voxel set to produce a third voxel set;

operating on the first voxel set and second voxel set to produce a fourth voxel set; and

operating on the third voxel set and fourth voxel set to produce a border for the region of interest.

13. The method of claim 12, wherein the operation of operating on the first voxel set and second voxel set to produce a third voxel set comprises:

determining a set union of the first and second voxel sets; and

inflating the set union to yield the third voxel set.

14. The method of claim 12, wherein the operation of operating on the first voxel set and second voxel set to produce a fourth voxel set comprises:

determining a set intersection of the first and second voxel sets; and

eroding the set intersection to yield the fourth voxel set.

10. A non-transitory computer-readable medium containing computer-executable instructions for determining a change in at least one portion of a biological object which, when executed by a hardware processing arrangement, configure the hardware processing arrangement to perform procedures comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the object;

determining a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image;

determining a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein;

normalizing the first region of interest;

determining the change in the at least one portion by comparing the normalized first region of interest to the second region of interest;

computing a first partial volume for the first region of interest;

computing a second partial volume for the second region of interest;

determining the change between the first and second partial volumes;

reducing a first noise associated with the first partial volume prior to determining the change between the first and second partial volumes; and

reducing a second noise associated with the second partial volume prior to determining the change between the first and second partial volumes.

11. A non-transitory computer-readable medium containing computer-executable instructions for determining a change in at least one portion of a biological object which, when executed by a hardware processing arrangement, configure the hardware processing arrangement to perform procedures comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the object;

determining a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image;

determining a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein;

normalizing the first region of interest by determining a normalization factor, and adjusting a signal associated with one of the first image or the second image by the normalization factor; and

determining the change in the at least one portion by comparing the normalized first region of interest to the second region of interest.

24. The computer-readable medium of claim 11, wherein, when the hardware processing arrangement executes the software instructions, the hardware processing arrangement is further configured to perform procedures comprising:

computing a first partial volume for each of a plurality of voxels associated with a first border of the first region of interest;

computing a second partial volume for each of a plurality of voxels associated with a second border of the second region of interest; and

further determining the change as a function of the first partial volume and the second partial volume.

25. The computer-readable medium of claim 24, wherein at least one of (i) the first partial volume is provided as a function of a first signal intensity of a first set of voxels associated with an interior of the first region of interest, or (ii) the second partial volume is provided as a function of a second signal intensity of a second set of voxels associated with an interior of the second region of interest.

26. The computer-readable medium of claim 25, wherein at least one of (i) the first signal intensity is provided as a function of a first average signal intensity of the first set of voxels, or (ii) the second signal intensity is provided as a function of an average signal intensity of the second set voxels.

27. The computer-readable medium of claim 26, wherein at least one of (i) the first signal intensity is provided as a further function of a contrast between the first signal intensity and a third signal intensity of a third set of voxels associated with a background, or (ii) the second signal intensity is provided as a further function of a contrast between the second signal intensity and the third signal intensity.

28. The computer-readable medium of claim 24, wherein the change is based on a difference between the first partial volume and the second partial volume.

15. An apparatus for determining a change in at least one portion of a biological object, comprising:

at least one hardware processing arrangement which is configured to;

obtain a first three-dimensional image and a second three-dimensional image of the object,

determine a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image,

determine a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein,

normalize the first region of interest by determining a normalization factor, and adjusting a signal associated with one of the first image or the second image by the normalization factor, and

determine the change in the at least one portion by comparing the normalized first region of interest to the second region of interest.

34. The apparatus of claim 15, wherein the hardware processing arrangement is further configured to:

compute a first partial volume for each of a plurality of voxels associated with a first border of the first region of interest, wherein the first partial volume is a function of a first signal intensity of a first set of voxels associated with an interior of the first region of interest, and wherein the first signal intensity is provided as a function of a first average signal intensity of the first set of voxels, and a contrast between the first signal intensity and a third signal intensity of a third set of voxels associated with a background;

compute a second partial volume for each of a plurality of voxels associated with a second border of the second region of interest, wherein the second partial volume is provided as a function of a second signal intensity of a second set of voxels associated with an interior of the second region of interest, and wherein the second signal intensity is provided as a function of an average signal intensity of the second set voxels, and a contrast between the second signal intensity and the third signal intensity; and

further determine the change as a function of the first partial volume and the second partial volume, wherein the change is based on a difference between the first partial volume and the second partial volume.

16. A method for determining a volume change between a first acquired three-dimensional image and a second acquired three-dimensional image associated with a biological object, comprising:

calculating a normalization factor for a first region of interest defined in the first acquired three-dimensional image, wherein the first region has first voxels therein;

adjusting a signal strength of the acquired three-dimensional second image by applying the normalization factor thereto;

computing a first partial volume for the first region of interest;

computing a second partial volume for a second region of interest defined in the second image, wherein the second region has second voxels therein; and

using a hardware processing arrangement, determining a change in volume between the first and second partial volumes,

wherein the operations performed on the first and second regions of interest are not performed on any other portion of the first or second images.

17. The method of claim 16, wherein the operation of calculating the normalization factor for the first region of interest comprises:

determining a first average signal intensity of an interior of the first region of interest;

determining a second average signal intensity of an interior of the second region of interest; and

setting the normalization factor to equal the first average signal intensity divided by the second average signal intensity.

18. The method of claim 17, wherein the second region of interest comprises a group of voxels, and the adjusting procedure comprises scaling a signal intensity for each voxel within the group of voxels by the normalization factor.

23. The method of claim 16, further comprising at least one of displaying or storing data associated with the change in volume in a storage arrangement in at least one of a user-accessible format or a user-readable format.

35. The method of claim 17, wherein (i) the computing the first partial volume comprises computing a partial volume for each of a plurality of voxels associated with a first border of the first region of interest as a function of the first average signal intensity and a contrast between the first average signal intensity and a further signal intensity of a further set of voxels associated with a background, (ii) the computing the second partial volume comprises computing a partial volume for each of a plurality of voxels associated with a second border of the second region of interest as a function of the second average signal intensity and a contrast between the second average signal intensity and the further signal intensity, and (iii) the change is based on a difference between the first partial volume and the second partial volume.

19. A method for determining a change in a hippocampus of a brain, comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the brain;

determining a first region of interest within at least one of the first and second three-dimensional images, the first region of interest including first voxels therein and encompassing the hippocampus;

determining a second region of interest within the other of the first and second three-dimensional images, the second region of interest including second voxels therein and encompassing the hippocampus;

normalizing the first region of interest by determining a normalization factor, and adjusting a signal associated with one of the first image or the second image by the normalization factor; and

determining the change in hippocampus by comparing the normalized first region of interest to the second region of interest.

20. The method according to claim 19, wherein the operations performed on the first and second regions of interest are not performed on any other portion of the first or second images.

21. A non-transitory computer-readable medium containing computer-executable instructions for determining a volume change between a first acquired three-dimensional image and a second acquired three-dimensional image associated with a biological object which, when executed by a hardware processing arrangement, configure the hardware processing arrangement to perform procedures comprising:

calculating a normalization factor for a first region of interest defined in the first acquired three-dimensional image, wherein the first region has first voxels therein;

adjusting a signal strength of the second acquired three-dimensional image by applying the normalization factor thereto;

computing a first partial volume for the first region of interest;

computing a second partial volume for a second region of interest defined in the second image, wherein the second region has second voxels therein; and

determining a change in volume between the first and second partial volumes, wherein the operations performed on the first and second regions of interest are not performed on any other portion of the first or second images.

29. The computer-readable medium of claim 21, wherein the computing the first partial volume comprises computing a partial volume for each of a plurality of voxels associated with a first border of the first region of interest, and the computing the second partial volume comprises computing a partial volume for each of a plurality of voxels associated with a second border of the second region of interest.

30. The computer-readable medium of claim 29, wherein at least one of (i) the first partial volume is provided as a function of a first signal intensity of a first set of voxels associated with an interior of the first region of interest, or (ii) the second partial volume is provided as a function of a second signal intensity of a second set of voxels associated with an interior of the second region of interest.

31. The computer-readable medium of claim 30, wherein at least one of (i) the first signal intensity is provided as a function of a first average signal intensity of the first set of voxels, or (ii) the second signal intensity is provided as a function of an average signal intensity of the second set voxels.

32. The computer-readable medium of claim 31, wherein at least one of (i) the first signal intensity is provided as a further function of a contrast between the first signal intensity and a third signal intensity of a third set of voxels associated with a background, or (ii) the second signal intensity is provided as a further function of a contrast between the second signal intensity and the third signal intensity.

33. The computer-readable medium of claim 29, wherein the change is based on a difference between the first partial volume and the second partial volume.

22. A non-transitory computer-readable medium containing computer-executable instructions for determining a change in a hippocampus of a brain which, when executed by a hardware processing arrangement, configure the hardware processing arrangement to perform procedures comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the brain;

determining a first region of interest within at least one of the first and second three-dimensional images, the first region of interest including first voxel therein and encompassing the hippocampus;

determining a second region of interest within the other of the first and second three-dimensional images, the second region of interest including second voxel therein and encompassing the hippocampus;

normalizing the first region of interest; and

determining the change in hippocampus by comparing the normalized first region of interest to the second region of interest.

36. A non-transitory computer-readable medium containing computer-executable instructions for determining a change in at least one portion of a biological object which, when executed by a hardware processing arrangement, configure the hardware processing arrangement to perform procedures comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the object;

determining a first voxel set comprising voxels associated with a first region of interest defined in the first image;

determining a second voxel set comprising voxels associated with a second region of interest defined in the second image;

operating on the first voxel set to obtain a first interior voxel set;

operating on the second voxel set to obtain a second interior voxel set;

operating on the first voxel set to obtain a first border voxel set;

operating on the second voxel set to obtain a second border voxel set;

computing a first partial volume for the first border voxel set as function of at least one of (i) a first signal intensity of the first interior voxel set, or (ii) a further signal intensity of voxels associated with a background;

computing a second partial volume for the second border voxel set as function of at least one of (i) a second signal intensity of the second interior voxel set, or (ii) the further signal intensity; and

determining a change between the first partial volume and the second partial volume.

37. The computer-readable medium of claim 36, wherein, when the hardware processing arrangement executes the software instructions, the hardware processing arrangement is further configured to perform procedures comprising:

calculating at least one of (i) a first normalization factor for the first region of interest, or (ii) a second normalization factor for the second region of interest; and

adjusting at least one of (i) a first signal strength of the first voxel set based on the first normalization factor, or (ii) a second signal strength of the second voxel set based on the second normalization factor.

38. The computer-readable medium of claim 36, wherein at least one of the first interior voxel set or the second interior voxel set is obtained by at least one of (i) determining a first set intersection of the first and second voxels sets, and eroding the first set intersection, or (ii) eroding the first and second voxels sets, and determining a second set intersection of the eroded first and second voxels sets.

39. The computer-readable medium of claim 36, wherein at least one of the first border voxel set or the second border voxel set is obtained by at least one of (i) determining a set union of the first and second voxel sets, (ii) inflating the set union, or (iii) determining a difference between the inflated set union and an interior set associated with at least one of the first voxel set or the second voxel set.

40. A method for determining a change in at least one portion of a biological object, comprising:

obtaining a first three-dimensional image and a second three-dimensional image of the object;

determining a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image;

determining a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein;

normalizing the first region of interest;

determining the change in the at least one portion by comparing the normalized first region of interest to the second region of interest;

using a computing arrangement, computing a first partial volume for the first region of interest;

computing a second partial volume for the second region of interest;

determining the change between the first and second partial volumes;

reducing a first noise associated with the first partial volume prior to determining the change between the first and second partial volumes; and

reducing a second noise associated with the second partial volume prior to determining the change between the first and second partial volumes.

41. A system for determining a change in at least one portion of a biological object, comprising:

at least one hardware processing arrangement which is configured to;

obtain a first three-dimensional image and a second three-dimensional image of the object;

determine a first region of interest within at least one of the first three-dimensional image or the second three-dimensional image;

determine a second region of interest within another one of the first three-dimensional image or the second three-dimensional image, wherein the first and second regions have voxels therein;

normalize the first region of interest;

determine the change in the at least one portion by comparing the normalized first region of interest to the second region of interest;

compute a first partial volume for the first region of interest;

compute a second partial volume for the second region of interest;

determine the change between the first and second partial volumes;

reduce a first noise associated with the first partial volume prior to determining the change between the first and second partial volumes; and

reduce a second noise associated with the second partial volume prior to determining the change between the first and second partial volumes.