Measuring biological tissue parameters using diffusion magnetic resonance imaging

• US 9,404,986 B2
• Filed: 05/07/2012
• Issued: 08/02/2016
• Est. Priority Date: 05/06/2011
• Status: Active Grant

• Current Assignee: Regents of the University of California (University of California)
• Original Assignee: Regents of the University of California (University of California)
• Inventors: White, Nathan S., Dale, Anders M.
• Primary Examiner(s): Arana, Louis
• Application Number:
  US13/466,081
• Publication Number:
  US 20120280686A1
• Time in Patent Office: 1,548 Days
• Field of Search: 324/309, 324/307, 324/306
• US Class Current: 1/1
• CPC Class Codes: G01R 33/48 NMR imaging systems
• CPC Class Codes: G01R 33/56341 Diffusion imaging

Abstract

Methods, devices and systems are disclosed for measuring biological tissue parameters using restriction spectrum magnetic resonance imaging. In one aspect, a method of characterizing a biological structure includes determining individual diffusion signals from magnetic resonance imaging (MRI) data in a set of MRI images that include diffusion weighting conditions (e.g., diffusion gradient directions, diffusion gradient strengths, sensitivity factors (b-values), or diffusion times), combining the individual diffusion signals to determine a processed diffusion signal corresponding to at least one location within one or more voxels of the MRI data, calculating one or more parameters from the processed diffusion signal by using the diffusion weighting conditions, and using the one or more parameters to identify a characteristic of the biological structure, in which the one or more parameters include values over a range of one or more diffusion length scales based on at least one of diffusion distance or diffusion rate.

Claims

1. A method of characterizing a biological structure, comprising:

determining individual diffusion signals from magnetic resonance imaging (MRI) data in a set of MRI images of a biological structure, the MRI data including diffusion weighting conditions that include at least one of diffusion gradient directions, diffusion gradient strengths, sensitivity factors (b-values), or diffusion times;

combining the individual diffusion signals to determine a processed diffusion signal corresponding to at least one location within one or more voxels of the MRI data;

calculating one or more parameters from the processed diffusion signal by using the diffusion weighting conditions; and

using the one or more parameters to identify a characteristic of the biological structure,

wherein the one or more parameters include values over a range of one or more diffusion length scales based on at least one of diffusion distance or diffusion rate.

2. The method of claim 1, wherein each of the individual diffusion signals includes an orientation and a volume fraction.

3. The method of claim 1, wherein the one or more diffusion length scales include fine scale diffusion processes occurring at one end and coarse scale diffusion processes occurring at the other end of each of the one or more diffusion length scales, wherein the fine scale diffusion processes include transverse diffusivities less than longitudinal diffusivities, and the coarse scale diffusion processes including transverse diffusivities greater than longitudinal diffusivities.

4. The method of claim 1, wherein the one or more parameters include at least one of scaled volume fraction, fiber orientation distribution, compartment geometry, geometric tortuosity, or compartment diameter.

5. The method of claim 4, wherein the fiber orientation distribution includes a hindered fiber orientation distribution (h-FOD) and restricted fiber orientation distribution (r-FOD), and the compartment geometry includes a cylindrical geometry.

6. The method of claim 5, further comprising producing a model of a microstructure of the biological structure based on the fiber orientation distribution.

7. The method of claim 6, wherein the model includes at least one of intra-cylindrical diffusivity values or extra-cylindrical diffusivity values of fibers within the biological structure.

8. The method of claim 7, wherein the compartment geometry further includes a cylindrical geometry.

9. The method of claim 8, wherein the model includes at least one of intra-spherical diffusivity values or extra-spherical diffusivity values of the spherical compartments within the biological structure.

10. The method of claim 8, further comprising deriving cellularity index values over the range of the one or more diffusion length scales for each voxel of the one or more voxels.

11. The method of claim 10, wherein the cellularity index values are derived based on maximizing sensitivity to the spherical compartments in the r-FOD.

12. The method of claim 5, further comprising producing a tractography model of fiber tracts of the biological structure by using fractional anisotropy values.

13. The method of claim 3, wherein the calculating includes fitting estimate values into a linear mixture model of the processed diffusion signal for a plurality of conditions of the diffusion weighting conditions.

14. The method of claim 13, wherein the fitting includes using at least one of spherical harmonic functions or Bessel functions.

15. The method of claim 1, wherein the biological structure includes at least one of cell bodies, axons, dendrites, cell processes, intracellular, or extracellular space.

16. The method of claim 1, wherein the characteristic of the biological structure includes at least one of a categorization as a healthy or a diseased structure, severity of the diseased structure, shape, size, density, or orientation.

17. The method of claim 1, wherein the values corresponding to the identified characteristic include a relative value compared to a normal value of the characteristic.

18. The method of claim 1, further comprising forming an image of the identified characteristic at a particular diffusion length scale using the value of at least one of the one or more parameters at the corresponding diffusion length scale.

19. The method of claim 18, further comprising forming a combined image of the identified characteristic from each of the formed images across the range of the one or more diffusion length scales.

20. The method of claim 13, further comprising acquiring the MRI data by using a bias variance tradeoff in the linear mixture model.

21. An MRI system, comprising:

a magnetic resonance imaging data acquisition system adapted to acquire magnetic resonance (MR) data from a biological structure of a subject, the MR data comprising one or more data voxels; and

a data processing unit that receives the MR data from the magnetic resonance data imaging acquisition system, the data processing unit comprising;

a mechanism that generates diffusion data of the biological structure from the MR data by determining a diffusion signal corresponding to at least one location within the one or more data voxels by combining a plurality of individual diffusion signals, each individual diffusion signal having an orientation and volume fraction, the diffusion data including one or more parameters, and

a mechanism that uses the one or more parameters to identify a characteristic of the biological structure,

wherein the one or more parameters include values over a range of one or more diffusion length scales based on at least one of diffusion distance or diffusion rate, and

wherein the MR data includes at least one of diffusion gradient directions, diffusion gradient strengths, diffusion waveforms, diffusion sensitivity factors (b-values), or diffusion times.

22. The MRI system of claim 21, wherein the one or more diffusion length scales include fine scale diffusion processes occurring at one end and coarse scale diffusion processes occurring at the other end of each of the one or more diffusion length scales, wherein the fine scale diffusion processes include transverse diffusivities less than longitudinal diffusivities, and the coarse scale diffusion processes including transverse diffusivities greater than longitudinal diffusivities.

23. The MRI system of claim 21, wherein the one or more parameters include at least one of volume fraction, fiber orientation distribution, compartment geometry, geometric tortuosity, or compartment diameter.

24. The MRI system of claim 23, wherein the fiber orientation distribution includes an h-FOD and an r-FOD, and the compartment geometry includes a cylindrical geometry.

25. The MRI system of claim 24, wherein the data processing unit further comprises a mechanism that produces a model of a microstructure of the biological structure based on the fiber orientation distribution, the model including at least one of intra-cylindrical diffusivity values or extra-cylindrical diffusivity values of fibers within the biological structure.

26. The MRI system of claim 25, wherein the compartment geometry further includes a cylindrical geometry, and the model further includes at least one of intra-spherical diffusivity values or extra-spherical diffusivity values of the spherical compartments within the biological structure.

27. The MRI system of claim 26, wherein the data processing unit further comprises a mechanism that derives cellularity index values over the range of the one or more diffusion length scales for each voxel of the one or more voxels, wherein the cellularity index values are derived based on maximizing sensitivity to the spherical compartments in the r-FOD.

28. The MRI system of claim 26, wherein the data processing unit further comprises a mechanism that produces a tractography model of fiber tracts of the biological structure using fractional anisotropy values.

29. The MRI system of claim 21, wherein the biological structure includes at least one of cell bodies, axons, dendrites, cell processes, intracellular, or extracellular space, and wherein the characteristic of the biological structure includes at least one of a categorization as a healthy or a diseased structure, severity of the diseased structure, shape, size, density, or orientation.

30. A computer program product comprising a nonvolatile computer-readable storage medium having instructions stored thereon, the instructions comprising:

code for determining individual diffusion signals from MRI data in a set of MRI images of a biological structure, the MRI data including diffusion weighting conditions that include at least one of diffusion gradient directions, diffusion gradient strengths, sensitivity factors (b-values), or diffusion times;

code for calculating one or more parameters of the MRI data by combining the individual diffusion signals to determine a processed diffusion signal corresponding to at least one location within one or more voxels of the MRI data, wherein the combining includes fitting estimate values into a linear mixture model of the processed diffusion signal for a plurality of conditions of the diffusion weighting conditions; and

code for using the one or more parameters to identify a characteristic of the biological structure.

31. The nonvolatile computer-readable storage medium of claim 30, wherein the one or more parameters include values over a range of one or more diffusion length scales based on at least one of diffusion distance or diffusion rate.

32. The nonvolatile computer-readable storage medium of claim 30, wherein the characteristic of the biological structure includes at least one of a categorization as a healthy or a diseased structure, severity of the diseased structure, shape, size, density, or orientation.

33. The nonvolatile computer-readable storage medium of claim 31, wherein the values corresponding to the identified characteristic include a relative value compared to a normal value of the characteristic.

34. The nonvolatile computer-readable storage medium of claim 31, further comprising code for forming an image of the identified characteristic at a particular diffusion length scale using the value of at least one of the one or more parameters at the corresponding diffusion length scale.

35. The nonvolatile computer-readable storage medium of claim 34, further comprising code for forming a combined image of the identified characteristic from each of the formed images across the range of the one or more diffusion length scales.

36. A method for estimating parameters of a linear mixture model of diffusion magnetic resonance imaging data, comprising:

receiving MR data from diffusion MRI signals of a biological structure;

using model regularization of a solution vector to estimate model parameters from the MRI data by placing constraints on the form and the extent of the estimated model parameters; and

using the estimated model parameters to identify a characteristic of the biological structure.

37. The method of claim 36, wherein the placing constraints includes imposing at least one of a smoothness, sparceness, or non-negative constraints on the solution vector.

38. The method of claim 36, wherein the constraints include prior assumptions and vary as a function of location within a voxel of one or more voxels of the MR data with respect to the biological structure.

39. The method of claim 36, further comprising remove particular contributions from the diffusion MRI signals while preserving desired contributions of the diffusion MRI signals to estimate other model parameters.

40. The method of claim 39, wherein the biological structure includes at least one of cell bodies, axons, dendrites, cell processes, intracellular, or extracellular space.

41. The method of claim 40, wherein the particular contributions include extracellular water diffusion signal that is associated with vasogenic or cytotoxic edema, and the desired contributions include grey and white matter diffusion signal, wherein the identified characteristic includes diffusion properties of grey and white matter, the diffusion properties including orientation, density, and diameter distribution of axons and dendrites.

42. The method of claim 36, wherein the estimated model parameters include isotropic restricted diffusion values within at least one of cylindrical or spherical geometries.

43. The method of claim 42, further comprising using the isotropic restricted diffusion values to identify the characteristic including at least one of a categorization as a healthy or a diseased structure, severity of the diseased structure, shape, size, density, or orientation.

44. The method of claim 42, further comprising using the estimated model parameters over multiple time points to quantify changes of the characteristic of the biological structure over time.

45. The method of claim 42, further comprising using changes in isotropic restricted diffusion to measure changes in tumor cellularity over time, and changes in tumor cellularity with or without treatment.